We limited participation in this shootout to UPS systems costing less than Rs
12,000–systems aimed at home and SOHO users, as well as individual office desktops.
The rationale behind this limit is that bar the corporate sector, users can’t afford
decent power conditioning after investing in mid to high-end systems.
We got 25 units for review at PCQ Labs in
April. One model was priced well over the Rs 12,000 limit. Another was faulty and did not
work during power testing. A third had to be dropped when its back-up electronics failed
consistently during testing. And two (the Tata Liebert and Computer Ware’s Compact)
came in over a week after our testing was over. Some vendors also submitted several units
that had only minute differences on their VA rating. In such cases, we included only one
of the supplied models in the ranking process.
After completeing our tests and analyzing the
results, we noticed a few interesting trends.
The Trends
The above-300 VA models performed better, both
due to their being somewhat better value with more reserve power available well within
their rated limits. If we kept all the 17 ranked UPSs in one list, the top five places
went to above-300 VA models.
The US brands did not fare very well, in
comparison to "homegrown" ones. In essence, they lost out in voltage
regulation–they were unable to hold their output to within 10 percent of 220 volts,
especially when the mains voltage surged high above or fell well below the rated limit.
These UPSs are ideal for situations where the mains supply is stable and well-regulated.
However, in many Indian situations, that is not true, and good regulation–with
undervoltage and overvoltage protection-- is very important.
Editors'' Choice
The overall ranking (see also the “The UPS Shootout” in this issue) lists all
the UPS units tested. For the Editor’s Choice awards, however, we classified the UPSs
into two groups separate groups–300 VA and below, and above 300 VA.
In the 300 VA and below category, our Editors''
Choice award went to the APC Backup Pro 280.
In the units above 300 VA category, our Editors’ Choice awards went to the Datex
Accure and the Nissan 625 M VAR.
Caveats
These reviews are meant to be a guide to UPS
selection, rather than a sweeping or generic endorsement.
What this means is this; Depending on your
individual needs and your local electrical conditions and requirements, your personal
ranking may differ from ours.
For instance, in your area, the mains voltage
may be quite stable, with the occasional minor or infrequent drop in line voltage. In this
case, even a UPS that does not correct very well for overvoltage may be sufficient. Or
your work place has a centralized power correction system, such as a CVT or a servo
stablizer. Again, in such cases the regulation requirements may be different (see
“Planning Your Power” in this issue).
And then there’s an important element
that we could not test fully in PCQ Labs: long-term reliability. We simulated events that
can cause failures, such as full output short-circuits, surges and drops in line voltage,
and short-term doubling of the rated load, In addition to these, many other factors,
including design, engineering, and component and overall assembly quality, affect
long-term reliability.
As a result, users often put the reputation of the brandname above other parameters, and
not without reason.
Finally, there is always the little chance of
the unit tested performing differently from the units that ship. This is especially likely
in devices such as UPSs, which involve analogue and power electronics whose quality can
vary.
The only such case we got to know about
involved a UPS in 1994, unfortunately the winner of an Editors’ Choice award. When we
later got reader complaints about the unit, a Quasar Microsine online, switching off
erratically under normal use, we bought a unit from the market and confirmed the problem.
So though we have had only one such report over the past three years, we now take a good
deal of care at PCQ Labs to inspect all product reviewed for signs of "souping
up", and make random checkswith existing users of the product or its predecessors.