How do uninterruptible power supplies work

How do uninterruptible power supplies work?

Before delving into the details of how an uninterruptible power supply works, let's take a quick look at the basics of this type of equipment. The fundamental purpose of a UPS is to provide an uninterruptible source of power for the equipment it protects. How exactly is this done? An electric device plugged into the wall (or into a surge suppressor plugged into the wall) has only one source of power. If there is a blackout, the electricity is cut and the device obviously goes off immediately. A UPS changes this equation by providing its equipment two sources of power.

UPSes are designed so that there is one source of power that is normally used, called the primary power source, and another source that kicks in if the primary is disrupted, called the secondary power source. The power from the wall is always one of these sources, and the battery contained within the UPS is the other. A switch is used to control which of these sources powers the equipment at any given time. The switch changes from the primary source to the secondary when it detects that the primary power has gone out. It switches back from the secondary power source to the primary when it detects that the primary power source has returned.

Online ("True") UPS

The online UPS, sometimes called a true UPS, is the best type you can buy. Paradoxically, it is both very similar to, and totally opposite to, the least-expensive type, the standby UPS. It is very similar to it in that it has the same two power sources, and a transfer switch that selects between them. It is the exact opposite from the standby UPS because it has reversed its sources: in the online UPS the primary power source is the UPS's battery, and utility power is the secondary power source!

Block schematic of an online ("true") UPS. You will notice if you look at the schematic for the standby UPS that it is identical, except that the primary and secondary power paths
have been exchanged, and here the battery is the primary power source.

Of course, while seeming small, this change is a very significant one. Under normal operation the online UPS is always running off the battery, using its inverter, while the line power runs the battery charger. For this reason, this type of UPS is sometimes also called a double-conversion or double-conversion online UPS. This design means that there is no transfer time in the event of a power failure--if the power goes out, the inverter (and its load) keeps chugging along and only the battery charger fails. A computer powered by an online UPS responds to a power failure in the same way that a plugged-in laptop PC does: it keeps running without interruption, and all that happens is that the battery starts to run down because there is no line power to charge it.

You may ask yourself, why bother having the secondary power path (the dashed line in the diagram above) if you are always running off the battery anyway? The reason is that this provides backup in the event that the inverter fails or stutters due to some sort of internal problem. While unusual, this can happen, and if it does, the unit will switch to the filtered, surge-suppressed line power. In this event, the matter of transfer time comes into play again, just as it does when a standby UPS reacts to a power failure. Of course, power failures are much more common than inverter failures.

There is another key advantage to having the equipment running off the battery most of the time: the double-conversion process totally isolates the output power from the input power. Any nasty surprises coming from the wall affect only the battery charger, and not the output loads.

Even though it may appear from the schematic diagrams that the online UPS and standby UPS have the same components inside, this is not the case. The distinction is that there is a big difference between designing chargers and inverters that are normally sitting around doing nothing and only run say once a month for a few minutes, and designing ones that are running 24 hours a day for weeks on end. The additional engineering and the increased size and quality of the components combine to make online UPSes much more expensive than lesser designs. They are typically used only for large servers, and for backing up multiple pieces of equipment in data centers. They are available in sizes from about 5,000 VA up to hundreds of thousands of VA and even larger.

Aside from the cost, a disadvantage of the online UPS is its inefficiency. All the power going to the loads is converted from AC to DC and back to AC, which means much of the power is dissipated as heat. Furthermore, this is happening all the time, not just during a power failure, and while running equipment that draws a lot of power. To combat this shortcoming, a new design called a delta-conversion online UPS was created. "Delta" is the scientific term often used to refer to the differential between two quantities. In this design, the battery charger is replaced with a delta converter. Instead of providing all of the output from the battery under normal circumstances, some of it is provided directly by the delta converter from the input line power. In the event of a power failure, the delta converter stops operating and the unit acts like a regular double-conversion online UPS, since the inverter is also running off the battery all the time.

Simplified block schematic of a delta-conversion online UPS. The converter and inverter both handle DC and AC current, providing the two power sources of the UPS. Bypassing the battery for part of the power during normal operation reduces power consumption.

This is a new design and is also available only in large UPSes (over 5,000 VA). They can result in substantial energy savings costs for large units.

Standby UPS / Standby Power Supply

The standby UPS is the simplest and least expensive UPS design. In fact, some don't even consider a standby UPS to really be a UPS, calling it instead a standby power supply (SPS). However, many of the most common consumer-grade devices marketed as UPSes, particularly on the lower end of the budget scale, in fact use this general design. They are sometimes also called offline UPSes to distinguish them from online UPSes.

In this type of UPS, the primary power source is line power from the utility, and the secondary power source is the battery. It is called a standby UPS because the battery and inverter are normally not supplying power to the equipment. The battery charger is using line power to charge the battery, and the battery and inverter are waiting "on standby" until they are needed. When the AC power goes out, the transfer switch changes to the secondary power source. When line power is restored, the UPS switches back.

Block schematic of a standby UPS. The primary power source is filtered and surge-suppressed to protect against line noise and other problems that would not cause a switch to battery power.

While the least desirable type of UPS, a standby unit is still a UPS and will serve well for most users. After all, if standby UPSes didn't work, they wouldn't sell. For a very critical function, however, such as an important server, they are not generally used. The issue with a standby UPS is that when the line power goes out, the switch to battery can take a few miliseconds, which can cause problems, while with continuous UPS there is no switchover time (as the name suggests!), making much more stable.