CTI 2572 ETHERNET TCP/IP ADAPTER MODULE TECHNICAL OVERVIEW Version 1.
Copyright 1993, 1998 Control Technology Inc. All rights reserved. This manual is published by Control Technology Inc., 5734 Middlebrook Pike, Knoxville, TN 37921. This manual contains references to brand and product names which are tradenames, trademarks, and/or registered trademarks of Control Technology Inc. and Siemens AG. Other references to brand and product names are tradenames, trademarks, and/or registered trademarks of their respective holders.
PREFACE This Technical Overview Manual provides an introduction to CTI 2572 Ethernet TCP/IP Adapter and to the networking specifications used with the module. The manual is intended for those individuals who wish to gain a fundamental understanding of the 2572, Ethernet, TCP/IP, and SIMATIC TI505 PLC command interfaces.
TABLE OF CONTENTS CHAPTER 1. INTRODUCTION.................................................................................................. 1 1.1 Description ........................................................................................................................... 1 1.2 2572 Functions ..................................................................................................................... 2 CHAPTER 2. LOCAL AREA NETWORK OVERVIEW .................................................
TABLE OF FIGURES Figure 1 CTI 2572.......................................................................................................................... 1 Figure 2 Token Ring Topology....................................................................................................... 4 Figure 3 Ethernet Topology (Coax) ............................................................................................... 5 Figure 4 10Base5 (Thick Coax) Topology ...............................................
CHAPTER 1. INTRODUCTION 1.1 Description The 2572 Ethernet TCP/IP Adapter is a single wide I/O module which connects SIMATIC TI505 controllers to Ethernet networks. Using TCP/IP (Transmission Control Protocol/ Internet Protocol), other controllers or computers can access registers, counters, timers, and memory locations in the PLC (Programmable Logic Controller).
1.2 2572 Functions Server Mode In Server Mode, the module responds to PLC commands embedded in the TCP/IP packet. The initiating device (usually a computer or a 2572 operating in client mode) encapsulates the PLC command in the TCP/IP protocol and transmits it to the IP address of the 2572. The 2572 retrieves the command from TCP/IP and formats a corresponding request to the PLC processor.
CHAPTER 2. LOCAL AREA NETWORK OVERVIEW Local Area Networks (LANs) grew from the need to interconnect many intelligent devices in a non-proprietary manner. Today, the most widely used LAN standards are Token-Ring (IEEE 802.5) and Ethernet (IEEE 802.3). Token Bus (IEEE 802.4), was proposed by GM in the late 80’s as a part of the Manufacturing Automation Protocol (MAP). Token Bus has seen limited use, primarily due to cost and installation difficulty.
2.2 IEEE 802.5 (Token-Ring) Token Ring (IEEE 802.5) uses a ring topology. Each station on the network has two connections - one to the "downstream" station and one to the "upstream" stations. Each station receives network data from the downstream station and passes it to the upstream station. The data circulates through the network until it reaches the initiating station, which validates that it was correctly sent though the network.
2.3 Ethernet and IEEE 802.3 Officially, "Ethernet" is the name of a CSMA/CD network developed at Xerox. It was jointly documented by Digital Equipment Corporation (DEC), Intel, and Xerox in a 1982 specification for Ethernet Version 2. This specification is commonly called Ethernet II or "Blue Book" Ethernet for the color of the specification cover. Later, the IEEE developed a specification for a CSMA/CD local area network based on Ethernet II. This specification, called 802.
The primary disadvantage of Ethernet is that the response time is not consistent under widely varying loads. As the network loading increases above 50% , collisions occur more frequently and the network response time increases. In practice, most large busy networks use bridges to reduce the number of collisions. Bridges are discussed in the next chapter. 2.4 IEEE 802.4 (Token Bus) Token-Bus is a LAN standard developed in conjunction with the General Motors Manufacturing Automation Protocol (MAP) effort.
2.6 Summary Older master/slave networks are rapidly being replaced by standard LANs which provide better performance, multi-vendor support, and integration with information networks. Of the current LAN standards, Token-Ring and Ethernet have the significant market share. Token Bus, because of the cost and complexity, has never become a major contender for the factory LAN market. In general, Ethernet offers the most cost-effective and flexible factory floor networking solution available today.
CHAPTER 3. ETHERNET OVERVIEW 3.1 ISO Reference Model Modern network implementations separate data communication functions into layers. Each layer performs a specific set of services and interacts with layers above and below in a well defined way. In 1984 the International Standards Organization published the Open Systems Interconnection (OSI) Reference model.
3.2 Ethernet Topologies and Media - Physical Layer Ethernet Version 2 and IEEE 802.3 are fully compatible at the physical level. Therefore the formal IEEE 802.3 specification is used to describe all Ethernet media. The 802.3 specification provides for a variety of topologies and media. All commonly used media operate at 10 Mb. Specified media/topology can be used interchangeably and can be intermixed on the same network.
Although 10Base2 cabling avoids the expense of a hub, it presents a major disadvantage for large installations. Because "T" connectors are used at each tap, there are two connections to the main cable for each node. Should a problem develop with the "T" connection, communications between stations above the connector and stations below the cable would be lost. In addition, if you need to add a station in the middle of a cabling run, you must splice in another connection.
FOIRL (Fiber Optic) Like the 10BaseT standard, the fiber optic cabling scheme uses a hub technology. Network stations are connected to the hub via point to point fiber optic connections. The fiber optic cable is attached to the network adapter using transceivers connected to the AUI port. A fiber optic cable run can extend up to 1 km.
Mixed Media The flexibility of cabling schemes and interconnection options is one of the strengths of the Ethernet physical layer. You may wish to consult your network vendor for additional details on the numerous options available. Fiber optic Backbone Fiber optic Backbone Fiber Optic Hub Fiber optic Backbone 10 BaseT Hub 10BaseT Hub - Multiple media types may be used on the same network.
The Ethernet Version 2 standard defines the packet as follows: Ethernet Destination Address Ethernet Source Address Type Data CRC 6 bytes 6 bytes 2 bytes Variable 4 bytes The type field is used to define the upper level protocol being used (e.g. TCP/IP, DEC LAT, etc.). The IEEE 802.
As you may notice, the Ethernet II format provides the Ethernet type information in 2 bytes while the 802.3 SNAP requires 8 bytes to provides this information. Consequently, most TCP/IP installations use the Ethernet II frame format. The 802.3 SNAP format is typically used only when strict 802.3 compatibility is required. 3.4 Ethernet Addresses Every Ethernet adapter board has a unique 48 bit physical assigned by the manufacturer of the adapter board.
CHAPTER 4. NETWORK PROTOCOLS 4.1 Connectionless vs. Connection Oriented Protocols In most networking discussions you will find references to connectionless and connectionoriented services. The following analogy may help you understand the features and advantages of each. Connectionless Protocols Connectionless services may be compared to the standard U.S. mail. You place a letter in an envelope and mail it.
Connection-Oriented Protocols Connection-oriented protocols may be compared to a personal telephone call. Before you begin a conversation, you establish a connection by dialing a number and verifying that you are talking to the desired person. Connection-oriented network protocols work in a similar manner. First, a logical connection is made between user applications.
IPX/SPX IPX (Internet Packet eXchange) and SPX (Sequenced Packet eXchange) are proprietary protocols used by Novell Netware. IPX is similar to Xerox’s Internet Datagram Protocol. IPX provides an unreliable connectionless transmission service. Since IPX is classified as an unreliable protocol, upper level command protocols (such as Netware Core Protocol) must provide their own delivery verification. The verification technique involves sending a single request packet and expecting an acknowledgement.
CHAPTER 5. TCP/IP OVERVIEW 5.1 TCP/IP Components TCP/IP consists of a hierarchy of related protocols. Upper level protocols are encapsulated in the data structure of lower level protocols. Figure 11 TCP/IP Data Encapsulation In a typical application using the 2572, the host computer application sends PLC commands via either TCP (Transmission Control Protocol) or UDP (User Datagram Protocol) services.
5.2 Internet Protocol Internet Protocol (IP) provides a basic packet delivery service. Technically, this service is defined as unreliable, connectionless, and best-effort. Unreliable means that delivery is not guaranteed by the protocol. The packet may be lost, duplicated, delayed, or delivered out of order. Connectionless, as defined previously, means that each packet is treated independently of all others.
A typical IP address consists of a network ID portion and a host ID portion. The distribution of the address bytes between network ID and host ID is defined by the address class. Class A uses the first byte for the network ID and the remaining three bytes for the host ID. Class B uses two bytes for the network ID and two bytes for the host ID. Class C uses the first three bytes for the network ID and the last byte for the host ID. The class type is determined by the first few bits of the address.
the datagram may actually pass through many gateways before reaching the destination network. IP has defined a comprehensive set of rules for processing datagrams. Among these specifications are the Internet Control Message Protocol (ICMP). Internet Control Message Protocol (ICMP) As described above, IP uses a best-effort method to deliver datagrams to their destination. There will be times IP cannot deliver a datagram to its destination (host system down, network congestion, etc.).
IP G UDP headers are contained in IP UDP IP DATA HEADER datagrams. Data is directed to a HEADER particular process by the use of a port number. In essence, a port is a message queue for the destination SOURCE DESTINATION LENGTH CHECKSUM application. Certain port numbers PORT PORT may be universally assigned and are UDP called well known ports. Echo, G Time, Nameserver, and SNMP (Simple Network Management Protocol) use well known ports.
TCP will notify both application programs. Typically one host computer (server) issues a passive open to a port, making the services available. Another host computer (client) issues an active open to initiate the connection. When the application session is complete, the connection is terminated by issuing a close to the port. TCP supports multiple connections to a single port and full duplex connections (two concurrent data streams flowing in the opposite direction).
5.6 Berkeley Sockets Although TCP/IP defines the network data structures and the rules for processing network traffic, it does not define exactly how application software interfaces with TCP/IP protocols. This attribute has given developers the freedom to optimize software for various processor class, operating systems, and programming languages.
CHAPTER 6. PLC COMMAND INTERFACE 6.1 Non Intelligent Terminal Protocol As discussed in previous sections of this manual, application programs communicate with the PLC by sending messages to the 2572 encapsulated in the TCP/IP protocol. The most common message format used by SIMATIC TI505 series PLCs is the Non Intelligent Terminal Protocol.
6.2 Task Codes As noted above the first word of the body of an NITP message contains a command. SIMATIC TI505 PLCs use a command structure known as task codes. Task codes are a set of numeric command codes which allow an external device to supervise and control the operation of the PLC. For example, using task codes a device can retrieve data from PLC memory.