UPS Architectures: Challenges and solutions

Data is critical to modern life and we are creating more of it than ever before. What’s more, thanks to the Internet-of-Things, a growing volume of today’s data is being created and stored without any direct human involvement. Given our reliance on it, not only do we need data to be stored safely and securely, we need 24/7/265 access - from anywhere in the world.

Most data is stored within, or backed up to, data centers that include layers of redundancy so that the failure of storage media such as a disc drive cannot result in lost information. However, loss of power to the servers and storage media in these centers can result in corruption, loss of service and an impact on data availability. To minimize these possibilities, data centers rely on Uninterruptible Power Supplies (UPS) to ensure that power is constantly available or, in the event of a very long outage, a controlled shutdown is possible.

There are many different types of UPS, offering different levels of performance, size and costs. What they all share is the use of an energy store – most often a battery – to provide power during mains supply disturbances or blackouts. Three of the most popular UPS types include:-

Offline (Standby) UPS: In this configuration, the load is fed directly from the mains input meaning that the energy storage components are not in use for most of the time. In the event of a mains failure, the static switch connects the inverter output to the critical load. This can introduce a ‘glitch’ of around 2-10ms, which is acceptable for all but the most sensitive systems.

As offline systems use mains power directly while it is present, they are the most efficient solution and often the lowest cost. The major drawback is that they normally offer little or no power conditioning.

Line-interactive UPS: While basically an offline topology a ferroresonant or buck-boost transformer arrangement is added to regulate the raw mains supply. Ferroresonant designs also offer tight voltage regulation and can store enough energy to overcome the ‘glitch’ while the static switch turns the inverter on.

Online UPS: With this type, power is always drawn through the power conditioning circuitry during normal operation and fault conditions. These UPS types are often known as ‘double conversion’ as the incoming mains is rectified to DC, passed through the battery and then inverted back to AC for the load. They are slightly less efficient than offline types and can be more expensive. However, the power conditioning is much better and there is no ‘glitch’ in the event of a fault. For this reason, the majority of UPS in use today are online types.

Toshiba has created a white paper looking at UPS technology in general and specifically some of the latest advances in semiconductors and topologies that enhance the all-important efficiency of these systems. To download the white paper, please click here: