Abstract: Merging units are interface devices for electronic current and voltage transformers. To a certain extent, merging units realize the sharing and digitization of process-level data. As a data source for bay-level and station control-level equipment in digital substations conforming to the IEC 61850 standard, they play a crucial role. With the promotion and engineering construction of digital substation automation technology, the functional and performance requirements for merging units are becoming increasingly stringent. This paper introduces a new type of digital-analog dual-function merging unit based on the MPC827X, which effectively meets the actual needs of current digital substation construction and has received positive feedback in engineering projects. Keywords: Digital substation, Merging unit, MPC827X, FPGA Introduction The International Electrotechnical Commission (IEC) defines a device called a Merging Unit (MU) in the IEC 61850 and IEC 60044-7/8 standards. It serves as an intermediate interface between current and voltage transformers and protection and control devices, synchronously acquiring current and voltage signals and outputting digital information to secondary protection and control equipment. There are two output methods: one is the serial communication method based on Manchester encoding defined in IEC60044-7/8[1][2], and the other is the Ethernet communication method based on IEEE802.3 defined in IEC61850-9[3]. Since the former has a relatively slow transmission rate, which limits the sampling rate, the most common method in engineering implementation is the Ethernet communication method. Electronic current transformers (ECT) and EVT (Electronic Voltage Transformers) can be roughly divided into passive and active types in principle. The former is based on the Faraday magneto-optical effect and the Pockels electro-optic effect, while the latter uses Rogowski coils and capacitor voltage division. At present, the application of passive EVT has not been widely promoted due to technical reasons, while passive ECT has been engineered. Meanwhile, in the construction of digital substations, both traditional electromagnetic instrument transformers and electronic instrument transformers coexist. Electronic instrument transformers output serial digital signals, while traditional electromagnetic instrument transformers output analog signals. Therefore, two types of merging units have emerged: digital MU and analog MU. However, with the accelerating pace of digital substation upgrades and construction, higher demands are being placed on both project cost and equipment performance, urgently requiring a merging unit with superior performance and more flexible configuration. The two types of merging units are shown in Figure 1. Please visit: Power Transmission and Distribution Equipment Network for more information. Figure 1-a Analog MU Figure 1-b Digital MU 1 Functional Module The definition of a merging unit was first given in IEC 60044-7/8. Generally speaking, a merging unit should have the following functional modules: (1) Electronic transformer interface: According to the definition of the merging unit in IEC60044-7/8, it needs to have 12 serial port input interfaces to receive the three-phase protection current, three-phase measurement current, three-phase voltage, neutral point current, neutral point voltage and bus voltage output by the electronic transformer. In order to meet the requirements of IEC61850-9 and in combination with the actual needs of the project, the merging unit introduced in this article can be expanded to have up to 16 serial input interfaces. (2) Sampling synchronization function: After correctly identifying the external input synchronization second pulse, a synchronization sampling signal is generated. The synchronization signal generally uses the second pulse provided by GPS. According to IEC60044-7/8, the rated standard values ​​of the frequency of the synchronization sampling signal are: 200f, 80f, 48f, 20f, f (for power frequency signals, f=50Hz). (3) Analog quantity acquisition and processing module: The analog MU needs to acquire multiple analog quantities and convert them into digital quantities for the CPU to read. (4) Sample value merging, output module: Generally, a 100BASE-FX Ethernet fiber optic interface is used to improve anti-interference capability, and the data packet frame format is defined in IEC61850-9. (5) Others: Meets a large number of other application requirements in digital substation automation engineering, such as data storage, human-machine interface, switch input and output, etc. 2 System structure Source: Power Transmission and Distribution Equipment Network Figure 2 shows the system structure of the merging unit, which has the following characteristics. 2.1 Ethernet communication platform based on MPC827X As the data source of protection and measurement and control equipment, the merging unit mainly outputs data through Ethernet to send data packets that conform to the IEC61850-9 frame format. Therefore, network performance is one of the most important performance indicators of the merging unit. MPC827X is a series of high-performance communication processors launched by Freescale after MPC860 with an excellent performance-price ratio. Its various technical features and technical indicators are very suitable for application in the automation intelligent devices of digital substations. Figure 2 Block diagram of the merging unit system based on MPC827X. MPC827X Power QUICC II is a dual-core architecture [4]: ​​(1) The embedded PowerPC core G2_LE core (a type of MPC603e) mainly handles high-level tasks, with 16KB instruction cache and 16KB data cache, and a maximum operating frequency of 400MHz; (2) The RISC-based communication processing module CPM (Communications Processor Module) is specifically responsible for the management of external communication interfaces, and its maximum operating frequency is 200MHz. The G2 core and CPM work together through the embedded high-speed dual-port RAM. This dual-processor architecture of MPC827X effectively reduces power consumption and improves data processing capabilities. The MPC827X boasts outstanding network communication capabilities. The CPM features two FCCs (Fast Communication Controllers) and four SCCs (Serial Communications Controllers). The FCCs can be configured as 10/100-Mbit Ethernet interfaces based on IEEE 802.3, and the SCCs can also be configured as 10-Mbit Ethernet interfaces based on IEEE 802.3. In addition to its superior network performance, the MPC827X also offers a comprehensive range of peripheral interfaces, such as I2C, SPI, USB, PCI, and UART, which can meet the diverse application needs of embedded systems for smart devices. Among these interfaces, the PCI interface is particularly noteworthy. The MPC827X's PCI bridge controller supports the PCI 2.2 bus standard, with a data width of 32 bits and a bus frequency of up to 66MHz, which provides the system with excellent peripheral expansion capabilities and driver software compatibility. The application of the PCI bus has several significant advantages for embedded systems: (1) Abundant peripherals. A wide variety of peripherals with PCI bus interfaces are available, such as Ethernet, USB, and serial ports, all of which are supported by mainstream chip manufacturers. (2) Fast access speed. The maximum data transfer rate of the MPC827X PCI bus is 266MB/s, and the data exchange between the PCI controller and the CPU adopts the DMA method. These characteristics enable the MPC827X PCI bus data throughput to meet the various needs of device products in digital substations. (3) Good software support. Because the PCI bus interface is widely used and highly standardized, operating systems such as Vxworks, Linux, and Windows have good software support for various mainstream PCI bus devices, providing a large number of standardized driver software, which greatly shortens the product development cycle. The merging unit discussed in this paper makes full use of the PCI bus expansion capability of the MPC827X: First, it uses Intel's high-performance Ethernet controller i8255x to expand 1-3 100M Ethernet ports; second, it uses a multi-serial port chip with a PCI interface to expand up to 16 serial ports for use as interfaces for electronic current transformers. In addition, the MPC827X also has a powerful and flexible MMU (Memory Management Unit), which can be well compatible with various bus-based memories and peripherals. In particular, the UPM (User-Programmable Machines) mode allows for highly flexible configuration of the timing of bus read/write signals and data to meet the needs of various bus-based external devices. Figure 3 shows the write command timing of a certain NAND Flash. Figure 3 NAND Flash Write Command Timing. In addition, the MPC827X also has an FPU (Floating-Point Unit) that supports hardware floating-point operations. 2.2 Multi-channel Ethernet Output Digital Substation Engineering Practice requires the merging unit, as the data source, to simultaneously provide data to multiple bay-level devices. Therefore, the merging unit is required to have multiple Ethernet output interfaces. The MPC827X has two FCCs, and with the addition of the PCI bus-extended Ethernet interfaces, it can have up to five. If more interfaces are needed, a switching chip can be used for expansion. This merging unit uses a multi-port Ethernet switching chip from MARVELL, which can expand to 6-12 100BASE-FX Ethernet interfaces. Each port is divided into a separate VLAN, so there is no mutual interference between ports. Figure 4 illustrates the scheme of using an Ethernet switching chip to expand ports. Source: www.tede.cn Figure 4 MPC827X Ethernet Port Expansion Scheme 2.3 FPGA Application Field Programmable Gate Array (FPGA) realizes the software-based hardware, with low data processing latency, stable and reliable operation, and convenient upgrades, making it a reasonable choice for implementing the synchronization module in the merging unit. The FPGA's inputs are the 1PPS and IRIG-B format time codes provided by GPS, and the outputs are the sampled pulses synchronized with 1PPS according to the IEC60044-7/8 standard and the decoded time stamp data. The synchronous sampling pulse triggers the electronic current transformer and the analog acquisition module of the merging unit to synchronously acquire current and voltage values. The decoded time stamp is read by the CPU through the serial communication port. At the same time, the FPGA needs to determine the validity of the 1PPS signal and eliminate false signals generated by 1PPS during fiber optic insertion/removal or other interference to prevent malfunctions of subsequent equipment. The FPGA notifies the CPU of the 1PPS signal validity determination result, and the CPU sets the corresponding flag in the IEC61850-9 message. The FPGA uses a high-precision crystal oscillator and maintains its own 1PPS clock. When the 1PPS provided by GPS is valid, its own 1PPS is synchronized; when the external 1PPS fails, the FPGA has a backup 1PPS, so the validity of data such as overcurrent protection and measurement and control can still be maintained. 2.4 High-precision ADC As the source of substation protection and measurement and control data, the merging unit places higher demands on the accuracy of the ADC. The currently widely used 14-bit ADC is already somewhat inadequate in the merging unit. In order to better adapt to higher application requirements, it is necessary to use a higher-precision 16-bit ADC. The merging unit uses two 8-channel 16-bit high-performance ADCs, which synchronously acquire and convert 16 analog signals for CPU reading under the trigger of the synchronous sampling signal. 2.5 Flexible configuration Because the actual needs of specific projects will vary, the merging unit must also be able to meet diverse application requirements to demonstrate its vitality in engineering applications. The electronic transformer interface, Ethernet interface, analog signal acquisition, human-machine interface, and digital input/output functional units of the merging unit all adopt a modular design, allowing for flexible configuration. For example, the instrument transformer interface can be configured with 16 analog inputs, 16 analog inputs plus 8 digital inputs, or 8 or 16 digital inputs; the Ethernet output module can be configured with 6 or 12 IEC61850-9-1 outputs, 4 or 8 IEC61850-9-1 outputs plus 2 GOOSE (Generic Object Oriented Substation Event) outputs; the human-machine interface (HMI) allows for optional switch inputs and outputs. The software also features a modular design, allowing for adaptation to different hardware module configurations by changing configuration files. 3. Conclusion Currently, digital substation automation systems using electronic instrument transformers and the IEC61850 standard are a research hotspot in power systems. This paper explores the implementation and technical characteristics of a new high-performance, highly flexible merging unit based on PowerPC. This merging unit has withstood practical testing in the construction and renovation of several digital substations in Beijing, Guangzhou, and other cities, achieving good social and economic benefits. References [1] IEC60044-7, Instrument Transformers-part7: Electronic Voltage Transducers[S], 1999. [2] IEC60044-8, Instrument Transformers-part8: Electrical Current Transducers[S], 2002. [3] IEC61850-9-1, Communication networks and systems in substations-Part 9-1: Specific Communication Service Mapping (SCSM) - Sampled values ​​over serial unidirectional multidrop point to point link, 2003. [4] MPC8272 PowerQUICC IITM Family Reference Manual, Rev.2, 2005.