5 Best Ways to Extend Ethernet

For many applications of TCP/IP networking, the 100-meter range limit of copper Ethernet cable becomes a problem. With conversion and extension, networks can cover distances that are measured in kilometers.

By Mike Fahrion August 3, 2012

When TCP/IP networking moves out of the home and office environment and into the real world, the 100-meter range limitation of copper Ethernet cable becomes a problem. The remote sensors along an oil pipeline, for example, are going to require a bit more range than that. Fortunately, there are ways to provide it. With conversion and extension, networks can cover distances that are measured in kilometers rather than meters.

1) Ethernet extender

One easy solution is the Ethernet extender. An Ethernet extender uses DSL technology to extend range up to 1,900 meters, or nearly 2 kilometers (Figure 1). And they’ll work with all sorts of copper wire, which makes them very cost-effective. The labor and cabling involved in a network installation normally represent a significant part of the expense, but Ethernet extenders will make use of any legacy cable already in place, like Cat5 cable or even old telephone lines. The savings can be substantial.

Ethernet extenders are set up in pairs. The first extender converts the Ethernet data for transmission over DSL. The second converts it back again. Throughput may be up to 50 Mbps, depending upon the range. Data signals attenuate over copper wire, of course, but even at the full 1,900 meters, Ethernet extenders can still provide throughputs of 1 Mbps. Some Ethernet extenders are also equipped to provide Power over Ethernet (PoE), which overcomes the challenge of powering remote devices placed away from a power source by providing power directly from the extended Ethernet port (Figure 2).

For best performance, be sure that the extender wire is free of load coils, filters, and splitters. 

2 and 3) Fiber optics: single-mode, multi-mode

When 1,900 meters isn’t enough, or enormous bandwidth is required, larger networks will typically employ fiber optic cable. There are two kinds. The cheaper option, multi-mode fiber optic cable, uses LED light and can carry data several kilometers. It is often used as the backbone infrastructure for office building and factory networks.

The telephone and cable companies, with their need for great range, are more likely to use single-mode fiber optic cable. Single-mode cable transmits with a laser rather than an LED, and it lets the telephone companies (telcos) transmit data across continents. Just as multi-mode fiber and its associated equipment is more expensive than copper wire, single-mode fiber optic installations are more costly than multi-mode. But many network designers still specify the more expensive single-mode fiber optics at the outset, reasoning that labor costs will represent a large part of the installation expense in any case, and that the great bandwidth and range provided by single-mode fiber ensure that the installation won’t become obsolete at any time in the foreseeable future (Figure 3). 

4) Radio

Copper or fiber optic installations may not make economic sense when data has to cross barriers like rivers or highways. But advances in radio technology are letting Wi-Fi do the job.

Radio waves can be absorbed or reflected by anything from the walls of a building to the vehicles in the parking lot. Called multipath propagation, the phenomenon means that transmitted signals arrive at the receiver at different times and out of sequence. The higher the radio frequency, the worse the problem becomes. Low frequencies, on the other hand, provide less bandwidth and need larger antennas and more power to produce useful gain. As a result, the most popular of the unlicensed frequencies was traditionally the 2.4 GHz band, which lies somewhere in the middle. But the 2.4 GHz band is used by everything from cordless telephones to microwave ovens, and has gotten very crowded.

Wi-Fi developers have responded to these issues with the new 802.11n standard. 802.11n Wi-Fi incorporates multiple-input multiple-output (MIMO) technology that uses multiple antennas at both the transmitting and receiving sides of the wireless connection. The data is split into numerous spatial streams, which are then transmitted through separate antennas and collected by corresponding antennas in the receiving devices. Onboard software uses signal processing algorithms to correct and interpret the incoming data.

MIMO 802.11n devices also use precoding and postcoding techniques like spatial beamforming. Spatial beamforming modifies the phase and relative amplitude of the signal to create a pattern of constructive and destructive interference in the wavefront, which simplifies interpretation on the receiving side. The 802.11n standard also adds frame aggregation to the MAC layer. By grouping several data frames into a single, larger frame, the ratio of payload data to total data volume is higher, as management information is specified less frequently. This allows for improved throughput. The 802.11n standard also adds 40 MHz channels to the physical layer (PHY), twice the bandwidth that was available under the older 20 MHz standard.

Using the IEEE 802.11n standard, a wireless network bridge (Figure 4) can create high-speed (up to 150 Mbps) point-to-point links between two remote devices or networks over line of sight distances of up to 15 km.

5) Cellular networking

Virtually everyone has a cell phone these days, and you could say that machines are starting to acquire them as well. Talk time has flattened out, and the wireless carriers know that cellular M2M networking will have to be one of their strategic growth markets. As a result, cellular data plans are becoming very M2M-friendly. Both Verizon and AT&T have recently announced shared-data plans (see “AT&T to introduce shared-data plans”) that are designed to encourage the networking of devices like industrial control systems, water meters, and ATMs.

When the cellular network serves as your Internet backbone, Ethernet’s original 100-meter range seems rather quaint. Using technology from companies like Czech Republic-based Conel, which connects machines through cellular networks to Ethernet, serial, and other data networks, users are creating networks of virtually unlimited size, with increasingly complex topologies. The systems support multiple protocols and encryption, the network remains under the user’s own control, and any station on the network can serve as an end or retranslation station. The use of multiple SIM cards in one device establishes redundancy, as the device is equipped to connect to more than one cellular provider at the same time. Cellular networking extends the network’s edge to any location that has cellular service.

Extension risks

Fiber optics are immune to electromagnetic interference (EMI), spikes, surges, and ground loops. The data isn’t traveling along a copper wire; it’s carried by a beam of light. This is invaluable in industrial applications, for example, where the electric motors on the machinery can generate powerful magnetic fields. Wi-Fi and cellular networking confer similar advantages.

But remember that extending your data communications range via copper wire increases the risk for unwanted electrical events. The greater the distance between connected devices, the more likely it is that they will have different building ground references and the associated risk for ground loops. If the cable is installed in an industrial environment and passing machinery along the way, greater range also creates more opportunities for EMI.

Copper wire network installations should be protected with surge suppression and isolation. Surge suppressors limit spikes between the signal and ground line and should be deployed as a first line of defense on power supply lines. Current models can be DIN rail mounted or connected directly to a cabinet, with surge protection ratings of up to 39 kA and less than 1 ns response time. But when the ground line rises, as it does in ground loop events, you’ll need isolation. Isolators convert data signals either to pulses of light or an electrical field, and then back again. Spikes and surges are stopped at the isolation zone. Isolators protect power lines by transforming dc power to ac, then back again.

Ethernet cable still has an effective range of 100 meters, just as it always has. But when used in combination with the technologies described above, there’s no reason you can’t extend your network to include just about any device, just about anywhere.

– Mike Fahrion, the director of product management at B&B Electronics, is an expert in data communications with 20 years of design and application experience. He oversees development of the company’s rugged M2M connectivity solutions for wireless and wired networks based on serial, Ethernet, wireless, and USB communication technologies. Fahrion is a speaker and author who writes a self-described politically incorrect newsletter, “eConnections,” with more than 50,000 monthly subscribers. Edited by Mark T. Hoske, content manager CFE Media, Control Engineering, Plant Engineering, and Consulting-Specifying Engineer, mhoske@cfemedia.com.


Control Engineering wireless coverage page has more about wireless networking. For more articles about system integration and industrial communications see https://www.controleng.com/integration.