In order to remain competitive, considerable time and money is being spent by all major cable operators to upgrade their networks and carry more than one-way video. Most traditional coaxial networks do not provide the reliability necessary for such services as telephony and high-speed Internet access. Tomorrow’s services can only be effectively delivered if many existing cable systems are upgraded. A key part of this upgrade is the choice in powering architecture and operators must think about the future. Upgrading an existing plant is expensive, and operators cannot afford to strand their investment. Engineering and technical managers must make decisions that will not only benefit their network today, but provide for a smooth and cost-effective migration for future network powering demands.

When planning the powering architecture, operators must take into consideration the following elements now and in the future:

• competition,
• types of services offered over the CATV network,
• density and demographics of the areas being served, and, of course,
• economics and budget limitations.

Powering Architectures
There are three basic powering architectures that can be implemented in a CATV network:

• Type #1: Distributed power without battery back-up (non-standby power supplies)
• Type #2: Distributed power with battery back-up (standby power supplies)
• Type #3: Centralized power with extended run time capability (powernode)

Each scenario has its own advantages and disadvantages. As demand on the network increases, the operator may migrate upward from one type of powering architecture to another. Think of each type as a step on a ladder, each rung requiring the network to achieve a higher level of reliability. Before an operator begins that climb, however, they must carefully consider the type of power supply they are purchasing and the technical capabilities of their power supply vendor to provide a cost-effective migration path.

Think of the Future
Reliability is of great concern to all operators, but the level of reliability built into a network can vary. For example, an operator running a video-only CATV network may not want to invest in standby power unless the CATV franchise agreement requires it. The operator could purchase standby power supplies without an inverter or batteries, saving approximately 50% of the power supply’s cost. Depending on volumes, an inverter and three batteries for a standby power supply could cost between $700-$900. An operator with 1,000 miles of coaxial plant could defer approximately one million dollars of capital investment in power supplies (based on one power supply per mile) by withholding deployment of the standby portion until the next “rung” of network reliability is required. However, the operator could lose valuable customer credibility and a competitive edge resulting from network downtime during short-term utility outages. In addition, if enhanced services are being considered, standby power will then be required for increased reliability. Since the future in any business plan is an unknown to some extent, it is wise to purchase power supplies that have the flexibility to grow with changing needs. In this age of razor-thin profit margins and high investment in upgrading plant, few, if any, operators can afford to strand their investment.

Even though the above scenario is possible, the majority of cable operators have some type of power protection and back-up, usually in a distributed setting. With this scenario, many power protection units are placed throughout a network. Each unit has its own set of batteries providing approximately one hour of back-up time. The upfront cost is higher than compared to a network without standby power protection, but reliability dramatically increases since the network can ride through most power outages. Enhanced services can now be delivered with a higher degree of network reliability.

Taking the next step up the ladder, from distributed to centralized, takes a little more thought and planning.

Making the Leap from Distributed to Centralized
Expansion of cable and telephony services requires two-way network capability plus a dramatic increase in reliability. As the future networks expand into these new service offerings, powering requirements go beyond the capabilities of traditional CATV powering architectures. The back-up power requirements for telephony and cable will have stretched the need for standby power anywhere from four to eight hours. To provide back-up power for extended periods of time, an alternate power source, typically a generator, must be used. For example, lifeline support, like the 911 telephone service, is a 24 hour-a-day requirement. Therefore, the cable network must be capable of operating through a prolonged power outage. Acceptable downtime, per Bellcore standards, is only 53 minutes per year which translates into a 99.999% network availability requirement, the same as provided by today’s telephone companies.

Not only is increased back-up time necessary, network powered loads such as telephones, network interface units and energy management controllers of the future turn off and on at different times and at different points on the network, creating varying powering demands. In addition, today’s interaction of constant power coax amplifiers increases the dynamic power demand on network power supplies. A centralized powering approach provides the architecture needed for greater reliability and stability. Operators are also competitively positioned within the industry, able to use reliability as a selling point and hopefully diffuse some of the negative perceptions that have long existed within the CATV market. Demographics and density of a service node also plays an important role in whether to adopt centralized powering. If the serving area is demographically attractive (i.e. income, education, multi-unit dwelling vs. single family, business vs. residential) and hook rates are generally high for enhanced services, then centralized powering should be considered. Subscribers willing to pay for services beyond video will demand uninterruptible service.

Economically, centralized powering can lower operator installation and service costs. One larger unit housing all node powering equipment is much easier to monitor and maintain than several smaller units scattered throughout a network, thus reducing mean time between repairs. Reliability is also dramatically increased. A centralized power supply usually has a generator for virtually unlimited run time or can support a larger number of batteries if a generator is not feasible. In comparison, smaller distributed units usually offer about an hour of battery back-up time with the standard three battery configuration. Centralized powering nodes also provide the added capability of N+1 redundant power modules, further increasing network reliability.

Growing Pains Don’t Have to Hurt
To help plan for power migration from one ladder rung to the next, operators should install power supplies capable of operating effectively in all three powering scenarios. Without a flexible “building block,” the operator will strand their original investment and incur additional and unnecessary costs. Looking at actual figures, a centralized powering unit typically costs between $15,000-$20,000 depending on power capacity, number of batteries and accessory equipment. If an operator is already using distributed power supplies designed for immediate redeployment into a centralized setting, operators can save $5,000-$7,000 per powernode.

However, special models are needed to make the migration smooth. An operator should not assume that all distributed standby power protection units can be readily deployed into a centralized setting. Several differences can exist between standby power supply units built for distributed and centralized settings.

In some cases, cable lengths for battery and coaxial hook-ups vary between distributed and centralized power supplies. Incorrect lengths can be a costly and time-consuming field upgrade. In other cases, the connectors needed for output voltage and battery cables are different between the two power supply designs. Retrofitting connectors is another unnecessary expense. Centralized power supply units typically use a generator for extended run time. The standby power supplies used in the distributed architecture need to be generator compatible. If not, an operator runs the risk of stranding their entire distributed power supply investment if the migration into centralized powering is undertaken. The power supply should have status monitoring capabilities as well. Constant communication with the power source greatly increases reliability through real time monitoring and proactive field maintenance practices.

Accu-Wave Recycling & Recovery Division