How Automotive Networks are ConfiguredPosted 1/17/2006
By Jared Busen, Lawrence E. Jeralds and Ben Komnick
Networks are an essential part of our lives. Internet, business communications and bank transactions are just a few of the many uses of networking. A local area network (LAN) is a network that is contained in one area, such as in a building or a vehicle. Two commonly used LANs are ethernet and controller area network (CAN). Ethernet is used extensively for home LANs and the Internet, while CAN is soon to be used for automotive LANs. In this article, we will look at some of the similarities and differences between the two local area networks: CAN and ethernet.
Many people are surprised to hear that vehicles on the road today contain a LAN. The automotive industry has used LANs for many years. One advantage to using a LAN on a car is that it eliminates point-to-point wiring. Point-to-point wiring means a sensor must be hardwired to every controller that needs its information, some of which could be on the far side of the vehicle. Vehicle networking eliminates redundant wiring because a sensor only has to be wired to the nearest controller, and the controller will send the information over the network.
When LANs were first introduced on vehicles, each manufacturer used its own proprietary protocol, a standard for communication. An automotive shop had to own several different scan tools to communicate with different makes of vehicles. The Society of Automotive Engineers (SAE) now requires that for diagnostic purposes, all light-duty vehicles use the same protocol for communication. The intent is that the scan tools from one manufacturer is capable of communicating with vehicles from other manufacturers. In 2008, SAE is requiring all light-duty vehicles to begin using CAN protocol, which SAE is calling standard J2284. CAN, developed by Bosch in 1982, is already in use in industries ranging from industrial machine control, medical equipment, aviation electronics and automated manufacturing equipment.
Ethernet is an established networking protocol used in an estimated 85 percent to 90 percent of all computer networks. Ethernet was developed by Xerox in the mid-1970s and is used extensively for the Internet and for school and home networks. If you have ever browsed the Internet, shared an electronic file, or purchased an item online, you have most likely used Ethernet.
Speeds and Topology
There are two speeds for CAN operation: CAN B for low speed and CAN C for high speed. CAN B is rated to transmit data at 83.3 kilobits per second (kbps) and is typically used for the interior communication requirements of the car. High-speed communications are not needed for things such as audio controls, comfort features and instrument panel displays. Topology is how the controllers are physically wired together. CAN B typically has a single wire bus topology (Figure 1) and is also configurable for a two-wire twisted pair bus (Figure 2).
CAN C, rated to transmit data at 500 kbps, is used for communication with the powertrain and suspension, where real-time high speed communication is necessary. High-speed information is needed for things such as stability control, antilock brake systems (ABS), air bags or active suspension. CAN C is configured for two-wire twisted pair bus topology.
A vehicle can have either low- or high-speed CAN or a combination of low- and high-speed CAN. If a vehicle has both low and high speeds, it needs a gateway to communicate between the two networks. A gateway is a device used to communicate between two different protocols, configurations or network speeds. In Figure 3, the gateway controller is the body control module (BCM).
Ethernet speeds range from 10 megabits per second (mbps) to 1 gigabit per second (gbps) and up to 40 gbps are expected in the future. Each speed requires a different cable for data transmission. Low speeds have used a coax or copper cable, medium speeds use twisted pair, and the highest speeds are using fiber optics.
"10Base2," "10Base5," "10 BaseT" or "100BaseT" are all terms used to describe ethernet. The first number designates the speed of the network; 10 stands for 10 mbps and 100 stands for 100 mbps. Base stands for using a baseband system, meaning a serial data system. The last digit stands for either the maximum length or type of cabling. A "2" means the maximum length is 200 meters, while a "5" would be 500 meters. If there is a "T" for the last digit, the cable used is a twisted pair.
The star topology (Figure 4) is the most popular type of topology, if you have a LAN in your house, it's probably a star. Star topology uses a central node that all the devices connect to - such as a hub, switch or router. Having a central node has numerous advantages. The central node acts as a repeater or signal amplifier, if needed. Since all data passes through the central node, it can be used for diagnostic purposes to manage or monitor data on the network. Some central nodes can be controlled remotely for maintenance and diagnostics. It is easy to add to the network by simply plugging a device into the central node. A central node can be chained to another central node, allowing for large expansion of the network. Lastly, the failure of one node will not affect the other nodes or cause the network to go down.
Bus topology (Figure 5) has a single cable called a main run cable to which all nodes are connected. At one end of the main run cable are terminators; a device that provides electrical resistance to absorb frames preventing them from being received again by the network. The CAN bus uses terminators at both ends of the main run cable or inside two of the controllers.
Because the devices on the bus do not need to run back to a central node, such as the star, there is less cabling involved to set up the network. A large problem with bus topology, CAN or ethernet, is that if one device fails, the devices after that could possibly lose their network access. Also, it's not as easy to expand the network with a bus topology as with a star topology. Instead of just plugging another computer into the node, you must physically tap into the main run cable.
If you've noticed the speed difference between CAN (500 kbps) and ethernet (up to 40 gbps) you may be wondering why there is such a large gap. The answer is actually quite simple - the speed of the computer. The controllers on the car process data at much slower speeds than the computer sitting at your desk.
Next month: Part two of the article will cover protocol and fault confinement.
AutoInc. Web Site |
ASA Web Site |
You Can Make a Difference in Automotive Repair Policy |
AutoInc's Annual Top 10 Automotive Web Sites |
OEM Web Sites: Overcoming Your Greatest Fears |
Hot and Ready: Vegetable Oil Used to Fuel Cars |
How Automotive Networks Are Configured |
Don't You Quit! |
Guest Editorial |
Tech to Tech |
Tech Tips |
News Briefs |
Taking the Hill |
Around ASA |
Shop Profile |
Net Worth |
Stat Corner |
Copyright (c) 1996-2011. Automotive Service Association®. All rights reserved.