-
Step 1: Decide what kind of backup you actually need
-
Step 2: Size the UPS against measured load, not the label
-
Step 3: Get the extension cord and surge protector rules right
-
Step 4: Pick battery chemistry for your runtime, not the spec sheet
-
Step 5: Run the install test everyone skips
-
Step 6: UPS won't turn on after an outage — work through this list
-
Where this checklist stops being useful
You're in one of three spots right now. Either you're speccing a UPS for a rack and want to get it right the first time, you're standing in front of one that won't power back on after an outage, or you're trying to figure out whether that extension cord feeding your surge protector is actually safe. All three are covered below. Six steps, roughly in the order you'll hit them. Jump to whatever applies.
Step 1: Decide what kind of backup you actually need
A UPS and an electric backup generator solve different problems, and people mix them up all the time. A UPS covers the gap between the outage and whatever comes next — your generator kicking in, or your equipment shutting down in an orderly way instead of just dying.
If your outages last hours and you've got critical loads, you probably need both. The generator replaces the wall outlet. The UPS covers the 10 to 30 seconds while the generator spins up and stabilizes its output. If your equipment can't ride through that gap — and a lot of networking gear can't — the generator alone isn't enough.
Here's the honest limitation on UPS: it buys time, not power. If you're routinely dealing with four-hour outages and you're expecting a UPS to carry you through, you're going to be disappointed no matter what capacity you buy. That's a generator or a large external battery bank problem, not a UPS problem.
Step 2: Size the UPS against measured load, not the label
When I first started speccing backup power, I assumed the VA number on the box was what I could actually run. It isn't. The rating is a ceiling, and you don't want to live at the ceiling.
Add up the nameplate wattage of everything you plan to plug in. Then derate. Motors, laser printers, and anything with a heating element pull a startup surge that can be several times their running draw. If you're sitting at 80% of the UPS's rated wattage with everything on, you're already too close — there's no headroom for the next thing someone plugs in.
This is where a lot of the confusion in reviews comes from. If you've been reading feedback on the CyberPower LX1500GU or other 1500VA line-interactive units, a common theme is runtime expectations, not hardware failures. People buy the VA number, plug in a workstation and a couple of monitors, and then find out they've got closer to eight minutes than the twenty they pictured. That's not the UPS underperforming. That's the load being higher than the buyer estimated.
Measure it. A $30 plug-in power meter tells you more in five minutes than the spec sheet will in an hour.
Step 3: Get the extension cord and surge protector rules right
This is the step most people skip, and it's the one that actually starts fires.
Can you plug an extension cord into a surge protector? Yes — the cord is just another load, and as long as the cord's own rating covers what you're plugging into it, that's fine. The problem is the other direction. Putting a surge protector on the end of an extension cord, or chaining two power strips together, is where things go wrong.
OSHA's flexible cord rules (29 CFR 1910.334) and NEC Article 400 both restrict flexible cords to temporary use rather than permanent wiring. Chaining them together turns every connection into a resistance point, and resistance under load means heat. That's the mechanical reason inspectors and manufacturers treat daisy-chained power strips as a violation in commercial spaces — not because the parts are defective, but because the setup has more failure points than anyone is going to check.
For a UPS specifically: plug it straight into the wall. Not into a surge protector, not into an extension cord. Most UPS manufacturers say the same thing, and it's not a warranty technicality. A surge protector upstream of the UPS can interfere with the transfer behavior, and it adds a failure point between your battery and the equipment that's supposed to stay up. (Worth mentioning: the UPS's own outlets are already doing surge suppression on most models — you're not gaining anything by putting another strip in front of it.)
Step 4: Pick battery chemistry for your runtime, not the spec sheet
Lithium UPS battery backup gets marketed hard right now, and some of that is justified. Lithium is lighter, handles more charge cycles, performs better in hot rooms, and comes with a battery management system that shuts things down before they go sideways. The honest limitation is cost per watt-hour.
If you need 45 minutes to an hour of runtime, lead-acid usually wins on price. Lithium wins when you're tight on rack space, when the battery gets cycled daily, or when the room runs hot enough that lead-acid would be losing capacity anyway. For a 5- to 10-minute ride-through in a network closet, the chemistry matters a lot less than people think — you're paying for cycle life you may never use.
One thing to check either way: the transfer time. Most line-interactive units switch in 4 to 10 milliseconds. If your equipment has a power supply that's picky about that gap — and some networking gear and medical devices are — you want an online/double-conversion unit instead. That's a topology question, not a battery question, and no amount of lithium fixes it.
Step 5: Run the install test everyone skips
Cold-start the UPS on battery before you trust it. Unplug it from the wall, press the power button, and confirm it actually comes up. Some units ship with the battery disconnected — there's a tab or a connector you have to engage before the first use, and it's easy to miss if you're installing at 11 p.m.
Then do a real outage test. Pull the plug from the wall — don't flip the breaker, because you want to test the UPS's transfer, not your panel. Watch what happens to your load. If anything reboots, you've found a problem before it found you.
Looking back on the first rack I signed off on, I should have tested runtime under actual load. I tested it with nothing plugged in and called it good. Three weeks later we had a real outage, and the runtime was a fraction of what I'd promised. The UPS was fine. My test wasn't.
Step 6: UPS won't turn on after an outage — work through this list
This one's for the situation where the power's back, but the UPS isn't. The list below is roughly the same regardless of brand — I've run it on CyberPower units, and I've run it on everything else.
- It's probably the battery, deeply discharged. After a long outage, the battery drains to the point where the UPS's internal logic won't let it start under load. Leave it plugged into the wall for 8 to 24 hours with everything turned off, then try again. This is the most common cause by a wide margin.
- Check for a resettable breaker on the back. Many units have a push-button circuit breaker that trips on overload or surge. If it's popped out, push it back in and try again.
- Try a cold start on battery. Unplug from the wall, hold the power button for a few seconds. If it starts on battery but not on line power, you're looking at an input-side problem — breaker, cord, or outlet.
- Consider battery age. Sealed lead-acid batteries in a UPS usually last three to five years. If yours is past that and the runtime has been shrinking, the battery may be worn out rather than discharged.
- Check the load. If something with a big inrush draw got plugged in during the outage, the UPS may be sitting in overload protection. Unplug everything and try starting it with no load.
- Check the room temperature. Some units have thermal cutoff. If the closet hit 100°F during the outage, give it time to cool down before you assume it's dead.
If none of that works, you're probably looking at a control board or transfer relay failure. That's a service call, not a DIY fix — and it's also a good argument for keeping one spare UPS sized for your critical load. A spare is cheaper than an emergency replacement at 2 a.m.
Where this checklist stops being useful
A few things that don't fit neatly into the steps above:
- Don't run a laser printer on a UPS. The fuser draws a surge on every print job that'll have the UPS beeping and possibly tripping into overload. Plug printers into a surge protector on the wall, not the UPS.
- A UPS is not a power strip. It has a limited number of outlets and a limited wattage for a reason. Adding a power strip to the back of a UPS to get more outlets defeats the purpose — now you've got more loads than the UPS was sized for, and no one's tracking the total.
- Put the battery replacement date on the unit. Sharpie on the side. When it hits year three, start budgeting. The failure mode for a UPS battery is usually graceful — runtime shrinks, then one day it doesn't hold at all — but 'graceful' still means 'during an outage.'
- Test under load at least once a year. Not just the self-test button. Actual load, actual outage simulation. The self-test checks the battery; it doesn't check whether the battery can carry what you've added since the last test.
This is accurate as of early 2025, based on current code references and hardware I've worked with. Battery pricing and model lineups move, so verify before you budget. And if you're speccing for a site with long outages, be honest with yourself about whether you're buying a UPS or a stopgap — those are different purchases.