You’re almost through with that important presentation that you will make to your client tomorrow. Suddenly, a power disturbance reboots your computer, and you loose all additions you made to the presentation since you last saved it. And what if the power glitch develops bad sectors in your hard drive, and you’re not able to open that presentation at all. It’s at times like these when you’ll miss having a good UPS. So what should you look for in a UPS.
Going by actual load considerations, a 1 kVA UPS should ideally be able to take the load of around 4–5 PCs with 15” color monitors at a power factor of .65. Most servers today, specify a higher power factor rating, so you may want to check with the vendor on whether his UPS can handle the load at this power factor. The power supplies of many PCs today are also being designed with better power factors, and in face, this rating is likely to improve further, and perhaps reach .9 or higher in the next two years or so.
Next comes backup time, which is determined by the number and rating of batteries used. First determine how much backup you need. If you have a generator backup, then around 10–15 mins should suffice. If you’re backing up mission critical servers, you may want longer backup times, of half hour or more. If the UPS unit doesn’t provide so much backup, ask for one where you can connect external batteries.
|Performance||Online UPS||LI UPS|
|Maximum output voltage||NA||20|
|Minimum output voltage||NA||20|
|Maximum switchover voltage||30||13|
|Minimum switchover voltage||30||13|
|Features (Both Online and LI)|
|Pricing (Both Online and LI)|
|For our final calculations, we used the sum of performance, features, and prices scores|
The type of batteries used in UPSs is also important. Most UPSs come with Sealed Maintenance Free (SMF) batteries. Just ensure that the vendor gives you warranty for both the batteries and the UPS, because the average life of UPS batteries is around 3 years.
Besides the load and backup time, there are several important factors in deciding a good UPS system. These can be broken up into electrical performance, the kind of features you want, and of course, the price. Check that your UPS has proper earthing, else its plug could give you a shock even when it’s not plugged in. In extreme cases, this may even send a bad spike down to the PC’s SMPS.
How we tested
We tested 16 reputed online and line interactive UPSs on the basis of their performance, features and price. The Brown-Gibson model was used for arriving at the weightages for all parameters.
Last year, we looked at 1 kVA UPSs priced less than Rs 15,000. This year, we removed the price barrier to accommodate the more expensive UPSs as well, and only asked for UPSs with internal batteries only. What we received was a mix of Online and Line Interactive UPSs, so we divided the shootout into these two categories. Online UPSs are more expensive, due to the technology. They offer a more stable output, and prevent most electrical disturbances and transients in the input from reaching the output load. That’s why they’re more suitable for protecting servers and other sensitive equipment. Line Interactive on the other hand, are less expensive, and more suitable for backing up small workgroups of a few computers. While they can also be used to backup servers, they may not be the best choice compared to the online counterparts. The various factors we considered in the evaluation are explained below.
Before running any tests, we first check the true rating of a UPS. Basically, if a UPS is rated at 1 kVA, we apply an equivalent load to it, and check whether it can sustain it for up to 30 secs. Only the UPSs that managed to sustain at least 80% of their rated load were ranked in the shootout. Interestingly, all the UPSs this time qualified, and were eligible for the remaining tests.
This gives us the maximum input voltage range over which a UPS doesn’t switchover to battery. The greater this range, the lesser the drain on battery. So a voltage range of 20% above or below 220 V was considered best.
This is the maximum and minimum output voltage that a UPS can supply. Ideally, the UPS should be able to keep its output to within 10–15% above or below 220V. We didn’t consider this parameter for online UPSs, because all of them provided a very stable output.
This is the characteristic that the UPS shows during switchover. A parameter called the Droop voltage measures the voltage drop experienced by a UPS during switchover to battery. This drop is measured across the computer SMPS, and is a DC voltage. This voltage has to be above 220 V. We also captured the switchover waveform on an oscilloscope to see the switchover time. This test was therefore not considered for online UPS systems.
Backup time for a UPS was measured at 80% of its rated load. Higher backup was better.
This is the DC voltage across the UPS’s battery terminals. When fully charged, this should be between 2.18 to 2.30 V DC for each cell. This gives an idea of the charger circuit of a UPS, and determines its battery life.
Here, we considered indicators, documentation, control software and special features like surge-protection port for telephone line, surge-protected power outlet for laser printers and external battery support. Control software is important as it can perform unattended system shutdown for 24-hour running load, like servers. UPSs that gave software as part of the package were given higher rating, followed by ones that gave it an option, and no points for those that didn’t have any at all.
Indicators are essential to monitor the health of a UPS, and know what’s going on. More indicators were given higher points. Finally, we felt that good documentation is important for a UPS, as it can help you install, and understand a UPS’s operation, and also help you troubleshoot it.
As always, we considered two factors under pricing—price and warranty. We considered the warranty for both the battery and the UPS.