I Almost Wrecked a $30K Battery Plant Commissioning in Kansas — All Because of One Power Supply Mistake

The Call That Started It All

It was a Tuesday afternoon in early October 2024 when Jackie called. Jackie is our contact at a battery plant in Kansas—a facility that reconditions industrial lead-acid cells. They were upgrading their entire communication system. New wireless handsets, a new PBX, the works. And they needed power supplies.

"We need something reliable," she said. "The old system would drop calls during charging cycles. The engineers say it's dirty power."

I nodded along. Sure, dirty power. I'd heard that before. "No problem, Jackie. I'll spec something solid."

Famous last words.

The Spec — What I Thought Was Fine

The plant had a rack of telecom equipment: a base station for the cordless phones, a router, a couple of small switches. Total draw? Maybe 12 amps at 24V. Nothing crazy. For the phone system itself, they needed a clean, regulated DC supply.

I grabbed a Mean Well NDR-240-24 off the shelf. Standard DIN rail unit. 240 watts, 10 amps, 24V. Good brand, right price point. Jackie had given me a budget—$250 for the power solution. The NDR series fit perfectly. I processed the order and moved on with my week.

A mistake that looked smart on paper. But I missed something critical.

The Disaster — September 2024, Welcoming Week

The plant scheduled commissioning for the week after Labor Day. Jackie's team installed everything. The new cordless phones—some mid-range Panasonic units—were powered on. The system booted. Phones registered. Calls went through.

Then the battery charging bays kicked in.

The phones started chirping. Not the ringing kind—the interference kind. Static on every call. Dropped connections. One handset completely rebooted mid-conversation. The foreman called it "the worst phone system since tin cans."

Jackie called me. "It's doing exactly what the old system did."

I felt my stomach drop.

The Investigation — What I Found

I flew out to Kansas the next day. Not my finest budget decision, but necessary. The plant is huge—rows of charging racks, forklifts buzzing around. The comms cabinet was tucked in a corner next to a main breaker panel. Classic industrial installation.

I put a multimeter on the output of the NDR-240-24. At idle: 24.1V, clean. During a charge cycle: 23.8V with a 120mV ripple spike. The scope showed it—nasty switching noise feeding back from the chargers.

The NDR series is a great general-purpose supply. But it's not designed for environments with heavy electrical noise. I'd chosen a cost-effective unit for a situation that demanded an industrial-grade filter.

What I needed? A Mean Well SDR-480-24. 480 watts, 20 amps, but more importantly, it has better ripple rejection and a higher tolerance for input transients. It's the unit the plant engineers should have seen on the spec sheet. But I never asked. I just assumed.

The Fix — And the Cost

I ordered three SDR-480-24 units overnight from Newark. Two for the comms cabinet (one primary, one backup), one spare. Cost: $189 each, plus $65 overnight shipping. Total: $632.

The NDR-240-24 I'd originally ordered? $147. The redo cost me nearly four times the original, plus a plane ticket ($380), plus a day of lost productivity.

Net screw-up cost: roughly $870 + a 2-day delay + a bruised relationship with Jackie.

I installed the SDR-480-24 the next morning. Fired up the call system. The phones stayed silent during charging cycles. Crystal clear calls. The foreman walked by and said, "This actually works."

Victory, but hollow.

The Lesson — Quality Is Brand Image

Here's what I learned, and it's a lesson that sticks with me every time I spec a supply now:

• You're not just buying a power supply—you're buying your client's trust. That NDR was a perfectly fine product. In an office, it would have been flawless. In an industrial facility with battery chargers and welding equipment? It was a liability.
• The $50 difference per unit would have saved me $800+ in rework. Penny wise, pound foolish—literally.
• Jackie's perception of Mean Well changed. Not because the brand failed, but because I failed to match the product to the application. She said, "I thought Mean Well was supposed to be good." It stung. Because it's true—the brand is good. I just used the wrong tool.

Per FTC guidelines (ftc.gov), advertising claims about reliability need substantiation. But there's no FTC rule for the trust you lose when you deliver the wrong spec. That's a different kind of liability.

What I Do Now — A Slightly Obsessive Checklist

After the Kansas incident—I call it "The Chirping Phone Disaster" in our team's internal wiki—I created a pre-order checklist. Every power supply order goes through it:

1. List all equipment on the circuit.
2. Check the ambient electrical environment (welding? chargers? VFDs?).
3. Verify ripple rejection specs against expected noise.
4. Confirm form factor fits the cabinet (DIN rail? chassis mount? open frame?).
5. Ask: is this a "good enough" choice or the right choice?

We've caught 47 potential errors using this checklist in the past 18 months. Not all were as dramatic as Kansas, but each one saved someone a headache.

A Final Rant on "Good Enough"

I'm not saying you always need the flagship unit. Mean Well's LRS series is perfectly fine for many applications. The SDR series is overkill for a clean office cabinet.

But here's the thing: the client's perception of your work is shaped by the first experience they have. If that first experience is a chirping phone during a battery charge cycle, they will remember that, not the invoice total. That $100 you saved becomes $10,000 in lost future business.

Honestly, I'm not sure why some engineers still treat power supplies as a commodity. My best guess is they've never dealt with the fallout of a bad spec. Lucky them.

Jackie and I still work together. She trusts us again—but it took a year to rebuild that confidence. And I still think about that Tuesday afternoon call. What if I'd just asked one more question?

A lesson learned the hard way. But learned.