At a glance: An Ethernet interface on a circuit board with headers for connecting to a CPU. Typical use: adding Ethernet to any microcontroller circuit. Ethernet support: 10BASE-T Source: EDTP Electronics (www.edtp.com) Hardware. The Packet Whacker (Network article 3-11) from EDTP Electronics is an Ethernet interface only. The circuit board contains a Realtek RTL8019AS Ethernet controller, an RJ-45 connector, two headers that bring out the signals required to communicate with the Ethernet controller, and related components. You can use the Packet Whacker to add Ethernet to just about any microcontroller. EDTP Electronics has similar boards with other Ethernet controllers: the NICki has a Cirrus 8900A and the NICkita has an ASIX 88796L. The Whacked 8051 Development Board includes Packet Whacker circuits and adds a Philips P89C668 8051-compatible microcontroller and 64 kilobytes of Flash memory. Software. EDTP provides example Packet Whacker firmware for UDP and TCP communications using a Microchip PIC16F877.

The Whacked 8051 Development Board includes C and Basic (BASCOM-51) code for UDP, TCP, and other Internet protocols. The NICkita includes an e-book with information about using the ASIX 88796L controller. Serial-to-Ethernet Bridge At a glance: enables RS-232 and RS-485 devices to communicate over networks Ethernet support: 10BASE-T Typical use: remote communications with devices with serial interfaces. Sources: Z-World (www.zworld.com), Netburner (www.netburner.com), R.E. Smith (www.rs485.com). Hardware. Thousands of existing devices that don’t support Ethernet have an RS-232 or RS-485 serial interface. With a serial-to-Ethernet bridge, you can communicate with these devices in an Ethernet network. The bridge connects to the device’s serial interface and to an Ethernet network.

Computers anywhere in the network can then exchange data with the device. After being configured, the bridge transparently sends received serial data on the network in TCP segments and sends data received in TCP segments to the device’s serial interface. Two devices that connect to a network via bridges can communicate the same as if they were connected directly by a serial interface. The bridge converts between interfaces as needed. Serial-to-Ethernet bridges are available from a variety of sources. Z-World’s EM1500 Multipoint Serial-to-Ethernet Bridge supports four RS-232 ports and one RS-485 port. The board’s CPU is a Rabbit 3000. R.E. Smith’s ESPSX3 Serial Port Server has two RS-232 ports and one port that is configurable as an RS-232 port or an isolated RS-485 port. The board contains a Rabbit Semiconductor RCM2200 module. Netburner’s SB72 Serial-to-Ethernet Device and Processor Board supports one RS-232 or RS-485 interface. The board’s CPU is a Motorola ColdFire 5272. Software. A Serial-to-Ethernet Server typically comes with an application that enables you to enter settings for your network and devices. Most also include a Web page that you can use for configuring when the device is on the network. After configuring, computers on the network can use TCP/IP applications to communicate with the device.

In Depth: Ethernet Controllers

An embedded system that supports Ethernet requires Ethernet controller hardware to provide the Ethernet interface. Many Ethernet controller chips are designed for use in desktop computers and include support for standard PC buses and Plug-and-Play functions. Small embedded systems typically don’t need all of the capabilities of a PC’s Ethernet controller. But because they’re available and familiar, a few of the older, simpler PC controllers have found new life in embedded systems. More recently, controllers designed specifically for use in embedded systems have become available. This section introduces some of the more popular controllers for embedded systems. If you buy a module with a controller on it, it’s likely that it will use one of the chips described below. A module containing an Ethernet controller will probably include firmware support for communicating with the controller. In many cases you can use the firmware without having to know much about the controller’s inner workings. Some vendors provide source code so you can customize if needed, while others release only the executable code. Even if you don’t need to program a controller directly, a basic understanding of how the controller works is helpful in selecting hardware and troubleshooting.

What the Hardware Does

Ethernet communications are typically handled by a combination of an Ethernet controller chip and device-driver code that communicates with the controller. Network article 3-12 shows the location of the Ethernet hardware and driver in a network stack. Many embedded systems use IP with TCP or UDP, but for some applications, the Ethernet driver can communicate directly with the application layer. The controller chip handles many of the details of sending and receiving Ethernet frames. In sending a frame, a controller typically does all of the following:

• Receives the message to send and the destination address from higher-level software.

• Calculates the Ethernet frame check sequence.

• Places data, addresses, and other information in the frame’s fields.

• Attempts to transmit the frame when the network is idle.

• Detects collisions, cancels any transmitted frame with a collision, and retries according to the protocol specified in the IEEE 802.3 standard (half-duplex interfaces only).

• Provides an indication of success or failure of a transmission. In receiving a frame, a controller typically does all of the following:

• Detects and synchronizes to new received frames.

• Ignores any frames that are less than the minimum size.

• Ignores any frames that don’t contain the interface’s address or a valid multicast or broadcast address in the Destination Address field.

• Calculates the frame-check-sequence value, compares the result with the received value, and indicates an error if they don’t match.

• Makes the received frame’s data and other information available to the receiving computer. Higher-level software reads the message and does whatever needs to be done with it.

Ethernet Controller Basics

In an Ethernet-capable embedded system, a CPU manages communications with the Ethernet controller. The minimum requirement for the CPU is a microcontroller with an external 8-bit data bus. Some of the controllers that have been popular in embedded systems were designed for use on expansion cards for the ISA bus of early PCs. An embedded system that uses an ISA-compatible controller can ignore any unneeded interrupt, address, and status and control pins. A shorthand term for a network interface controller is NIC. The same term can also refer to an expansion card that contains a network interface controller. Related Components A typical controller requires few additional components. For 10BASE-T and 100BASE-TX systems, the IEEE 802.3 standard requires an isolation transformer that also functions as a low-pass filter between the controller and the network’s RJ-45 connector. Filters that comply with the standard are readily available. Examples include the FA163079 from YCL Electronics and the PM-1006 from Premier Magnetics. The appropriate filters vary with the controller chip or the MAU or PHY that connects to the filter. The vendors of these components typically provide recommendations and advice in selecting filters. As explained in Chapter 2, many controller chips require few additional components to interface to twisted-pair cable, but fiber-optic or coaxial cable is likely to require additional MAU or PHY circuits. Other typical required components include a timing crystal to clock the controller chip and decoupling capacitors for the power pins. Some controllers also support an interface to a serial EEPROM, which can provide nonvolatile, read/write storage of configuration data such as the Ethernet hardware address. Most controllers also have status outputs for interfacing to LEDs.