It happened on a Wednesday morning, around 9:15. I remember the time because our sales rep texted me, "Is the network down?" before the server room's beeping even registered in my head. The lights flickered twice, then went off. The building got quiet. Somebody said "probably just a grid thing."
The power came back after 40 minutes—maybe 35, I wasn't timing. The network didn't come back with it. Our phone system rebooted into nowhere, the dispatch server was unreachable, and two account managers lost work they thought was autosaved. The bill for that morning, counting overtime and lost productivity, was around $14,000. Actually, when finance ran the numbers, it was $13,740. Still enough to hurt.
I should say up front: I'm a procurement manager, not an IT engineer. I've managed our infrastructure spend—roughly $140,000 a year now—for five years. My job is to make sure we don't waste money on brands, logos, or unnecessary features. That's why what I did next makes no sense, even to me.
I bought the cheap UPSs, obviously
In early 2023, we needed to replace UPS units in our main office and server closet. I compared prices from three vendors, like our policy says, and picked the lowest line items. Each no-name unit was about $230. The Tripp-Lite units from my predecessor's quotes were about twice that. I even wrote a little note in my approval request: "commodity hardware, same specs, 47% cheaper." It looked responsible. It wasn't.
I didn't think a UPS was complicated. A UPS is, at its core, a power supply inverter connected to a battery. Grid power goes out, the battery sends DC to the inverter, and the inverter makes AC for your equipment. Simple enough, right? That's the same mistake I made when I bought a cheap office chair and ended up paying for two: I confused a simple concept with a simple product.
Then the power went out for real
About three months later, we had a longer outage. It wasn't a storm. The utility had a maintenance issue. Whatever the cause, the network went down and stayed down. When power came back, five of the fourteen UPS units wouldn't turn on. One unit with a bad battery beeped and flashed its control panel lights in a pattern none of us had seen before. Three were just dead. The IT manager said, "They're all still under warranty." Great, except the "warranty" meant we'd have to send units back, wait a week, and pay expedited shipping while our servers sat unprotected.
In the same room, two older Tripp-Lite rackmount UPS units that I'd almost written off as "legacy gear" were humming, showing steady green control panel lights, and holding the network core through every second of the outage. One of my techs pointed at them and said, "I know you hate paying extra, but this is extra."
And here's the part that stuck with me: I had read forum complaints about "Tripp Lite UPS not turning on" and thought it was a brand problem. But our problem was the exact opposite—our Tripp-Lite units were the ones that turned on without hesitation. The units that wouldn't turn on were the cheap ones I'd chosen.
Comparing the two up close
When I looked at the two types side by side, I finally saw what I'd skipped on the spec sheet. The budget model had no replaceable battery pack. The Tripp-Lite rackmount model had a front panel that told you battery status, load level, and fault conditions. The budget units had vague indicator lights; the Tripp-Lite control panel lights had labels and a self-test button. The difference wasn't just brand polish. It was the ability to know what's wrong before the UPS fails completely.
We spent $68 to replace a dead battery in one of the old Tripp-Lite units. It was back in service the same day. The budget units? They went to an e-waste pile.
Why I started looking at TCO, not unit price
I'm still a cost guy. The Tripp-Lite rackmount UPS we eventually standardized on costs more than twice what the no-name units cost. But after that outage, the true cost of the "cheap" units was obvious:
We lost five units after one outage. If you divide the total downtime cost by the number of failed units, each cheap unit helped cause a lot more than $230 in damage.
We also had to run servers unprotected for days while waiting for warranty replacements. And we had to pay our own IT team overtime to babysit the reboot process during the next outage, because nobody trusted the remaining cheap units.
That's not a "premium brand" tax. That's the cost of false savings.
Now, when I pick UPS hardware, I use the same approach I use for any critical component: find the lowest total cost over the life of the equipment. Tripp-Lite's selector and calculator helped with that. I put in the actual load and our target runtime, and the tool suggested a 1500VA rackmount model with headroom. That was the first time any UPS vendor's tool made me realize my previous sizing method—"just get 1500VA, that should work"—was the equivalent of guessing.
What about a UPS that won't turn on?
If you search "Tripp Lite UPS not turning on," you'll find a lot of advice. Here's the part that matters: a UPS won't turn on if its battery is completely dead or disconnected. The unit may beep, the control panel lights may flash, and it might look broken—but the battery is usually the first suspect. Before you buy a new UPS, check the manual. Many Tripp-Lite models have user-replaceable battery packs. That's a lot cheaper than replacing the entire unit.
But don't grab a car battery charger and try to resurrect a sealed lead-acid UPS battery while it's connected to the unit. I had a technician who asked me, honestly, "how to use a battery charger?" after he'd hooked a 10-amp charger directly to a UPS battery. It didn't end well. The battery might charge a little, but it won't regain capacity, and you risk damaging the UPS logic board. The right approach is to disconnect the battery, use a smart charger that's designed for sealed lead-acid batteries, and only if the battery isn't old. If the battery is more than three years old, just replace it. That's a maintenance cost, not a product flaw.
I also learned the difference between a UPS and a battery charger, which sounds stupid but is actually the whole point. The inverter in a UPS is meant to power a load for the seconds or minutes it takes to shut down safely. It is not a charger. The battery in a UPS needs to be fresh enough to deliver that power when the grid disappears. The moment you buy a new UPS, schedule the battery for replacement in about three years—unless you enjoy watching the control panel lights blink through a self-test you can't fix. By the way, UL 1778 is the safety standard that governs UPS equipment in North America. It covers the inverter, battery, and transfer switching, but it doesn't cover a user who bypasses the unit's charging circuit.
Where we ended up
In 2024, we replaced every remaining budget unit with Tripp-Lite rackmount UPS units. It cost about $9,200. I know the number because I approved the PO. In the first six months after the upgrade, we went through two planned utility maintenance windows and one transformer fumble in our street. Every time, the network stayed up. The servers didn't crash. The control panel lights came on, ran whatever self-test they do, and stayed green. No one stayed late to reboot anything.
I have mixed feelings about calling any vendor "reliable" based on one year of experience. At least, that's been our experience so far—and it's a lot better than the alternative. The point isn't that I suddenly love spending money. It's that I understand what I'm buying. A UPS is a power supply inverter, a battery, and a transfer switch that only matters for a few seconds. When those seconds hit, price should be the last thing you're thinking about.
I know that sounds like a marketing line. I'm a procurement guy. I don't do marketing. I do math—and the math changed.
Leave a Reply