Network Power Supply Issues: Surge Suppressors and Uninterruptible Power Supply (UPS) Functions

Surge suppressors are usually mounted on a wall power socket, to which a networking device is connected. This type of surge suppressor has circuitry that is designed to prevent surges and spikes from damaging the networking device. A device called a metal oxide varistor (MOV) is most often used as this type of surge suppressor. An MOV protects the networking devices by redirecting excess voltages, that occur during spikes and surges, to a ground. Simply put, a varistor is a device that is capable of absorbing very large currents without damage. An MOV can hold voltage surges on a 120 V circuit to a level of approximately 330 V.

Unfortunately, an MOV may not be an effective means of protecting the networking device that is attached to it. This is because the ground also serves as the common reference point for data signals going into and out of the computer. Dumping excess voltages into the power line near the computer can create problems. While this type of voltage diversion can avoid damage to the power supply, it can still result in garbled data.

When surge suppressors that are located in close proximity to networking devices divert large voltages onto the common ground, this can create a large voltage differential between network devices. As a result, these devices can experience loss of data, or in some instances damaged circuits.

You should also be aware that this type of surge suppressor has a limited lifetime, dependent, in part, on heat and usage. For all of these reasons, this type of surge suppressor would not be the best choice for your network.

In order to avoid problems associated with surges, what you could do, instead of installing individual surge suppressors at each work station, is to use a commercial quality surge suppressor. These should be located at each power distribution panel, rather than in close proximity to the networking devices. By placing a commercial grade surge suppressor near the power panel, the impact on the network, of voltage surges and spikes diverted to ground, can be reduced.

The problem of sags and brownouts can best be addressed by the use of uninterruptible power supplies (UPS). The extent to which UPS must be provided for a LAN depends on factors such as the budget, the types of services the LAN provides, the frequency of regional power outages, and the typical length and duration of power outages, when they do occur. At a minimum, every network file server should have a source of backup power. If power wiring hubs are required, then they must also be supported with backup power. Finally, in extended star topology networks, where internetworking devices such as bridges and routers are used, power backup must be provided to them, as well, in order to avoid failures in the system. Where possible, power backup should also be provided for all work areas. As every network administrator knows, it does little good to have an operational server and wiring system, if they cannot ensure that computers will not go down before users can save their spreadsheets and word processing files.

Sags and brownouts are usually power outages that are of a relatively short duration, and are caused by something, such as a lightning strike. This creates a power overload, and trips a circuit breaker. Because circuit breakers are designed to automatically reset, they can work from the surrounding power grid to where the source of a short is located in order to re-establish power. This usually occurs within seconds or minutes.

Longer power outages can occur, however, when an event, such as a severe storm or flood, causes physical disruption of the power transmission system. Unlike shorter power outages, this type of disruption in service is usually dependent on service crews for repair.

An uninterruptible power source is designed to handle only short-duration power outages. If a LAN requires uninterrupted power, even during power outages that could last several hours, then a generator would be needed to supplement the backup provided by a UPS. Can you think of situations where LANs might need the added backup of a generator?

A UPS consists of batteries, a battery charger, and a power inverter. The functions of each are as follows:

• inverter - convert low-level direct current voltage of the batteries into the AC voltage, normally supplied by the power line, to networking devices
• battery charger - designed to keep the batteries in peak condition during periods when the power line system is functioning normally
• batteries - generally, the bigger the batteries in a UPS, the longer a period of time it will be able to support networking devices during power outages

A number of vendors have developed UPS systems. You will find that they differ in the following ways: the power storage capacity of the batteries; the power delivery capability of the inverter; and the operational scheme (whether they operate continuously, or only when the input voltage reaches a specific level). Also, the more features a UPS has, the more it costs.

As a rule, UPS devices that offer fewer features, and cost less money, are used as standby power systems only. This means that they monitor power lines. If, and when, a problem occurs, the UPS switches over to the inverter, which is powered by its batteries. The time needed for this switch to occur is called the transfer time. Usually, the transfer time lasts for only a short time. This does not usually present a problem for most modern computers, which are designed to coast on their own power supplies for at least a hundred milliseconds.

UPS devices that offer more features, and cost more money, typically operate online. This means that they constantly supply power from inverters, which are powered by their batteries. While they do this, their batteries continue charging from the power line. Because their inverters supply freshly generated AC, such UPS devices have the added benefit of ensuring that no spikes from the power line reach the networking devices that they serve. If, and when, the AC power line goes down, however, the UPS batteries will switch, smoothly, from recharging to providing power to the inverter. Consequently, this type of UPS effectively reduces the transfer time needed to zero.

Other UPS products fall into a hybrid category. While they appear to be online systems, they do not run their inverters all the time. Because of these differences, be sure to investigate the features of any UPS you plan to incorporate as part of a LAN installation.

In any event, a good UPS should be designed to communicate with the file server. This is important so that the file server can be warned to shut down files when the UPS battery power nears its end. Additionally, a good UPS reports instances when the server starts to run on battery power, and supplies this information to any work stations running on the network, after the power outage has occurred.

The focus of this chapter was network design and documentation.  You learned that:

• Layer 1 components include: plugs, cable, jacks, and patch panels
• to finish Layer 1 design, both a logical and a physical topology must be generated
• Layer 2 devices such as switches reduce congestion and collision domain size
• Layer 3 devices such as routers are used to build scalable internetworks (larger LANs, WANs, networks of networks), or to impose logical structure on the network
• databases, and other shared resources, as well as the LAN's link to WANs and to the Internet
• any time you install cable, it is important to document what you have done
• a wiring closet is a specially designed room used for wiring a data or voice network
• backbone cabling consists of the backbone cabling runs; intermediate and main cross-connects; mechanical terminations; and patch cords used for backbone-to-backbone cross-connection
• surge suppressors are an effective means of addressing the problems of surges and spikes

Now that you have worked through this chapter, you are ready to begin the structured cabling project which is covered in the next chapter.