A UPS PRIMER
UPS’ (Uninterruptible Power Supplies) are used to provide back up electrical power to critical components. In addition UPS systems contain components to protect and condition input power. There are numerous types and brands that supply varying quality, duration and security. The simplest type is the On Line system.
In this style the protected equipment is constantly operating off of battery power. The advantages of this system are that the electrical power is very clean and in the event of a power outage there is no switch over time problem.
The major disadvantage and reason that this type of system is discouraged is that there is no redundancy. If any component fails, the protected system is without power.
The most common type is the two mode UPS. The normal operating or stand-by condition is shown in the Blue path.
The components are running off of the input voltage. There are varying amounts of surge/sag suppression and filtering in this path. The transfer time for most standby UPS’ is in the 2-4 ms range, however some units have much longer transfer times
When the Transfer Switch is activated, the Red path shows the electrical flow. The switch can be activated by lack of source voltage, sags, surges, extreme noise etc.
This type of system can sometimes be used in a reverse mode. The Transfer Switch is reversed, meaning that the primary (normal) path is the Red path and the back up is the Blue path. This will give the advantages of the on-line system and still provide protection if a UPS component fails.
Another type of UPS system is the Ferro-resonant system. This type uses an iron core transformer that stores energy for the duration of the switch to battery power. Therefore there is no switchover time.
Since the output path is through a transformer in both primary and back-up modes, the power is very clean.
The disadvantages include output waveform distortions caused by the inductive characteristic of the iron core. This problem has been exacerbated by the increase of ‘Power Factor Corrected’ components, which draw only sinusoidal current from the supply.
The three mentioned configurations are the most common types of UPS systems, under 3kW. UPS systems can be sized from about 250 Volt Amps (VA) / 170 Watts (W) up to over 220,000W (220kW) in single or three phase sources. They can weigh anywhere from 20 lbs. to over 5500 lbs. The battery size and type will determine the run time on back up power.
SIZE DOES MATTER
Determining the size of a UPS system is dependent upon several factors.
Selecting Equipment to be Protected
Computers, drives and most peripheral equipment should be included on this list. This should include monitors unless an automatic command can be sent to save data and safely shut down the protected equipment. Components are usually supplied with a load rating. This can be on the outside of the unit or in the operator’s manual.
Battery run time
This is the maximum length of time the system will operate on battery/back-up power. This requirement can be calculated by the amount of time that it will take either to start up alternate power systems (such as a generator), or to properly save data and shut down the protected equipment.
CALCULATING POWER REQUIREMENTS
The power required by a component is commonly labeled in either Watts (W) or Volt Amps (VA). The Watt and VA levels can differ significantly (the VA rating will always be equal to or higher than the W rating). The ratio of W to VA is called the ‘Power Factor’. In a strictly resistive load such as a heater or a toaster, Watts and VA numbers are the same.
Some devices will use a ‘Power Factor Corrected’ power supply. These units will have a Power Factor almost equal to one (~0.99). These components typically include servers, hubs and storage systems. This means that a 1000VA rated component will have a 990W load.
A ‘capacitor input switch mode power supply’ (non-corrected) will have a power factor ratio of 0.55 to 0.75. This typically includes desktop computers and peripherals. This means that a 1000 VA rated component will have a 550-750W load. Electronic components generally have a Power Factor of 0.70.
RULE OF THUMB
There is no easy method of determining if a component has a ‘capacitor input switch mode’ or ‘power factor corrected’ power supply.
The capacity of UPS systems is given in both Watts and Volt Amps. Neither specification can be exceeded.
For instance, a UPS with a rated load of 1000VA and 800W will not support a drive array with a rating of 900VA and 900W (Power Factor = 1). The VA level will be sufficient, but the required Watts will overload the UPS.
Once you’ve calculated and specified a UPS that will properly handle a particular load. You can usually extend the battery run time to the required length by purchasing additional batteries.
OH BY THE WAY!!
An often overlooked phenomenon in these calculations is the start-up surge. When an electronic device is turned on, there are internal components that draw more current than in normal operation. The worst offender is a hard drive. It takes significantly more current to start the head plate spinning than to keep it in motion. This momentary (~100ms) draw can exceed the drive frame power specification by up to 5 times. A rack containing 96 individual drives can trigger a circuit breaker if they all spin up at the same time after a blackout.
Staggered startup, either manually or automatically will solve this problem.
Don’t forget to include enough source power to carry the load of the UPS system recharging after a power failure.
Conditioning & Protection
Normal electric company fluctuations can include;
A reduction of the AC voltage, at the power frequency, that lasts from a half cycle to a few seconds.
This is a transient electrical impulse with a duration much less than 1/2 cycle.
Surge / Overvoltage
A voltage that exceeds specified limits either for a short time (seconds) or longer periods.
These types of disturbances can occur several times per day throughout most of the country. When a UPS is in a normal mode (not on battery power) these disruptions will not generally cause the UPS to switch to battery mode. Some UPS manufacturers have adjustable threshold settings that can activate the back up at various levels of interference. To avoid excessive switchovers, a conditioning and protection sub-system should be used in the normal mode.
TESTING UPS SYSTEMS
In the real world the ‘Pull the Plug’ test is the only reasonable method of testing a UPS. As long as the user knows the limitations of this type test it is worth the effort.
A ‘Pull the Plug’ test for UPS systems will not fully test the UPS capability. It will test the load capability of the UPS.
During a real power failure there are numerous pieces of equipment, air conditioners, lights etc, on the building circuitry. Since the UPS input cable is still connected, these heavy loads can cause an effective short on the input cable. This can in some circumstances cause a delay in the UPS switchover response time.
Most large cities have consultants or re-sellers that can analyze site power. These comprehensive tests will give an analysis of any power-related problems and status of the UPS conditioning capabilities.
Avoiding problems usually starts with adhering to state and local electrical and building codes.
Hopefully this article will give you a start on understanding the requirements involved in selecting a UPS system. Don’t be afraid to ask if you have any questions.
Jim Alfonse, owner of Tri-Sys Designs, is a Systems Integrator with twenty-five years experience in the Broadcast Industry. He's designed, built and commissioned installations from Satellite News Vehicles to Production Suites to OB vans. Jim has been involved with several equipment manufacturers performing video standards compliance and signal integrity testing.
For more articles on various subjects visit www.tri-sysdesigns.com