Материал: m012900e

272 • ETHERNET Network communication

5.2.2.1ETHERNET

ETHERNET address (MAC-ID)

Each WAGO ETHERNET fieldbus coupler is provided from the factory with a unique and internationally unambiguous physical ETHERNET address, also referred to as MAC-ID (Media Access Control Identity). This can be used by the network operating system for addressing on a hardware level.

The address has a fixed length of 6 Bytes (48 Bit) and contains the address type, the manufacturer’s ID, and the serial number.

Examples for the MAC-ID of a WAGO ETHERNET fieldbus coupler (hexadecimal): 00H-30H-DEH-00H-00H-01H.

ETHERNET does not allow addressing of different networks.

If an ETHERNET network is to be connected to other networks, higherranking protocols have to be used.

Note

If you wish to connect one or more data networks, routers have to be used.

ETHERNET Packet

The datagrams exchanged on the transmission medium are called “ETHERNET packets” or just “packets”. Transmission is connectionless; i.e. the sender does not receive any feedback from the receiver. The data used is packed in an address information frame. The following figure shows the structure of such a packet.

Fig. 5-9: ETHERNET-Packet

The preamble serves as a synchronization between the transmitting station and the receiving station. The ETHERNET header contains the MAC addresses of the transmitter and the receiver, and a type field.

The type field is used to identify the following protocol by way of unambiguous coding (i.e. 0800hex = Internet Protocol).

5.2.2.2IP-Protocol

The Internet protocol divides datagrams into segments and is responsible for their transmission from one network subscriber to another. The stations involved may be connected to the same network or to different physical networks which are linked together by routers.

Routers are able to select various paths (network transmission paths) through connected networks, and bypass congestion and individual network failures. However, as individual paths may be selected which are shorter than other paths, datagrams may overtake each other, causing the sequence of the data packets to be incorrect.

Therefore, it is necessary to use a higher-level protocol, for example, TCP to guarantee correct transmission.

Modular I/O System

ETHERNET TCP/IP

ETHERNET • 273

Network communication

IP addresses

To allow communication over the network each fieldbus node requires a 32 bit Internet address (IP address).

Attention

Internet addresses have to be unique throughout the entire interconnected networks.

As shown below there are various address classes with net identification (net ID) and subscriber identification (subscriber ID) of varying lengths. The net ID defines the network in which the subscriber is located. The subscriber ID identifies a particular subscriber within this network.

Networks are divided into various network classes for addressing purposes:

•Class A: (Net-ID: Byte1, Host-ID: Byte2 - Byte4)

The highest bit in Class A networks is always 0. I.e. the highest byte can be in a range of

0 0000000 to 0 1111111.

Therefore, the address range of a Class A network in the first byte is always between 0 and 127.

•Class B: (Net-ID: Byte1 - Byte2, Host-ID: Byte3 - Byte4)

The highest bits in Class B networks are always 10. I.e. the highest byte can be in a range of

10 000000 to 10 111111.

Therefore, the address range of Class B networks in the first byte is always between 128 and 191.

•Class C: (Net-ID: Byte1 - Byte3, Host-ID: Byte4)

The highest bits in Class C networks are always 110. I.e. the highest byte can be in a range of

110 00000 to 110 11111.

Therefore, the address range of Class C networks in the first byte is always between 192 and 223.

Additional network classes (D, E) are only used for special tasks.

More information

i A more detailed description of these principles is found on the Internet under http://www.WuT.de/us_printmed.html (W&T, Manual TCP/IP-ETHERNET for Beginners).

Modular I/O System

ETHERNET TCP/IP

274 • ETHERNET Network communication

Each WAGO ETHERNET fieldbus coupler/controller can be easily assigned an IP address via the implemented BootP protocol. For small internal networks we recommend selecting a network address from Class C.

Attention

Never set all bits to equal 0 or 1 in one byte (byte = 0 or 255). These are reserved for special functions and may not be allocated. I.e. the address 10.0.10.10 may not be used due to the 0 in the second byte.

If a network is to be directly connected to the Internet, only registered, internationally unique IP addresses allocated by a central registration service may be used. These are available i.e. from InterNIC (International Network Information Center).

Attention

Direct connection to the Internet should only be performed by an authorized network administrator and is therefore not described in this manual.

Subnets

To allow routing within large networks a convention was introduced in the specification RFC 950. Part of the Internet address, the subscriber ID is divided up again into a subnetwork number and the station number of the node. With the aid of the network number it is possible to branch into internal subnetworks within the partial network, but the entire network is physically connected together. The size and position of the subnetwork ID are not defined; however, the size is dependent upon the number of subnets to be addressed and the number of subscribers per subnet.

Fig. 5-10: Class B address with field for subnetwork ID

Modular I/O System

ETHERNET TCP/IP

ETHERNET • 275

Network communication

Subnet mask

A subnet mask was introduced to encode the subnets in the Internet. This involves a bit mask, which is used to mask out or select specific bits of the IP address. The mask defines the subscriber ID bits used for subnet coding, which denote the ID of the subscriber. The entire IP address range theoretically lies between 0.0.0.0 and 255.255.255.255. Each 0 and 255 from the IP address range are reserved for the subnet mask.

The standard masks depending upon the respective network class are as follows:

•Class A Subnet mask:

•Class B Subnet mask:

•Class C Subnet mask:

Depending on the subnet division the subnet masks may, however, contain other values beyond 0 and 255, such as i.e. 255.255.255.128 or 255.255.255.248, etc.

Your network administrator allocates the subnet mask number to you. Together with the IP address, this number determines which network your PC and your node belongs to.

The recipient node, which is located on a subnet initially, calculates the correct network number from its own IP address and the subnet mask.

Only then does it check the node number and delivers the entire packet frame, if it corresponds.

Example of an IP address from a class B network:

Attention

Specify the network mask defined by the administrator in the same way as the IP address when installing the network protocol.

Modular I/O System

ETHERNET TCP/IP

276 • ETHERNET Network communication

Gateway

The subnets of the Internet are normally connected via gateways. The function of these gateways is to forward packets to other networks or subnets.

This means that in addition to the IP address and network mask for each network card it is necessary to specify the correct IP address of the standard gateway for a PC or fieldbus node connected to the Internet. You should also be able to obtain this IP address from your network administrator.

The IP function is limited to the local subnet if this address is not specified.

IP Packet

In addition to the data units to be transported, the IP data packets contain a range of address information and additional information in the packet header.

Fig. 5-11: IP packet

The most important information in the IP header is the IP address of the transmitter and the receiver and the transport protocol used.

5.2.2.3TCP protocol

As the layer above the Internet protocol, TCP (Transmission Control Protocol), guarantees the secure transport of data through the network.

TCP enables two subscribers to establish a connection for the duration of the data transmission. Communication takes place in full-duplex mode, i.e. transmission between two subscribers in both directions simultaneously.

TCP provides the transmitted message with a 16-bit checksum and each data packet with a sequence number.

The receiver checks that the packet has been correctly received on the basis of the checksum and then sets off the sequence number. The result is known as the acknowledgement number and is returned with the next self-sent packet as an acknowledgement.

This ensures that the lost TCP packets are detected and resent, if necessary, in the correct sequence.

TCP port numbers

TCP can, in addition to the IP address (network and subscriber address), respond to a specific application (service) on the addressed subscriber. For this the applications located on a subscriber, such as i.e. web server, FTP server and others are addressed via different port numbers. Well-known applications are assigned fixed ports to which each application can refer when a connection is built up.

A complete list of "standardized services" is contained in the RFC 1700 (1994) specifications.

Modular I/O System

ETHERNET TCP/IP