The Call That Almost Got Ignored
It was a Tuesday afternoon in Q2 2024 when I got the ticket: a customer saying their new Lithonia ceiling spotlight stopped working after three days. They'd attached a multimeter reading showing no DC output from the driver. Order value: $87. Three units: one eu2c emergency light, one LED ceiling spotlight, and a bag of chandelier parts for a retrofit they were doing at a small boutique inn outside of town.
My first reaction? Honestly, I almost flagged it as a typical user error. You know, wrong wiring, dimmer incompatibility, that kind of thing. We get complaints on maybe 0.3% of our annual 50,000-unit order volume, and maybe one in ten turns out to be a genuine product flaw. The rest are installation issues. But this guy had included a photo of his multimeter display — 0.04VDC on a driver rated for 24V constant current. That's not user error. That's a dead driver.
"The most frustrating part of these callouts: explaining to a non-technical buyer that their multimeter might not be lying. You'd think a zero reading would be obvious, but I've had distributors argue that their 'LED driver is fine, must be the fixture.'"
So I did something that our normal process doesn't always encourage for small orders: I pulled the batch records.
Chasing the Batch
That specific driver model — a common constant-current 700mA unit used in our ceiling spotlight series — came from a production run in March 2024. Lot code: 24C-788. We'd shipped 4,200 of them to various resellers and OEM customers. The customer's unit was serial number 1,877 out of that lot. I requested sample testing from our warehouse for that lot. We keep a 30-unit archive batch for each major lot for at least 18 months.
While waiting, I called the customer back. He was a small contractor doing his first commercial project — three rooms of a 12-room inn. He'd ordered the eu2c for exit-sign compliance, some chandelier parts for the lobby, and the spotlights for hallway accent lighting. His total order: $870 including shipping. A $30 LED driver had failed.
"I tested it with my own multimeter," he said. "I've been fixing old fluorescent luminaires for years, and a dead driver is a dead driver." He even offered to ship the defective unit back on his dime. Here's the thing: that level of diagnostic initiative is rare on small orders. Most people just return the whole fixture.
The quality lab got back to me two days later. They had tested six samples from the same lot. Pass rate: 100%. But I still wasn't satisfied, so I asked them to test the specific failure mode the customer described — constant current output under a resistive load simulating the LED array. That's when we found it: two out of ten units showed drift after 15 minutes of operation, dropping from 700mA to around 480mA. The symptom would only appear after thermal soak. The customer's driver had failed open entirely, but the underlying issue was a batch of under-spec capacitors from a new supplier.
The Cost of a Small Signal
This is where the causation reversal kicks in. People think that expensive brands have fewer defects because they use better components. The reality is: they catch more defects because they pay attention to small signals — like a $87 customer's multimeter photo. If that contractor had been a big-box reseller with a pallet of 500 units, our automated sampling might have caught that capacitor drift during extended burn-in. But on a small order, the first warning often comes from a real person with a $20 multimeter.
I still kick myself for almost dismissing it. If I'd followed the standard escalation procedure (RMA-only, no investigation under $100 order value), we'd have shipped him a new driver, he'd install it, maybe the same batch would fail again, and we'd lose a customer — and a safety-critical load in an emergency egress fixture. The eu2c wasn't involved in the failure, but the trust was.
We recalled the remaining units from that lot — about 3,900 still in distribution. Cost of recall: about $18,000 in replacements and freight. But that quality issue could have cost us a $22,000 redo if that contractor's customer (the inn owner) had filed a safety complaint. The capacitor supplier got replaced within the same quarter.
So, How to Test an LED Driver with a Multimeter (and Why You Should)
That experience also reminded me that small contractors often don't realize they can test drivers before installing them. Here's the quick version of what I taught that customer — and what I now include in our small-order welcome emails:
- Set your multimeter to DC voltage (200V range). LED drivers output a regulated current, but the voltage endpoint varies with load. Under no-load condition (disconnected LED), most constant-current drivers produce a high open-circuit voltage (typically 48-60V). If you see 0V, the driver is likely dead.
- Measure across the output terminals. Red probe to +, black to -. With the driver powered on, you should see something between 30V and 60V (depending on the driver spec). If it's zero or below 10V, it's failed open.
- If you have a resistor load, you can test current: select DC amps (10A range), break the circuit in series with the resistor, and read the current. It should match the driver's rated current (e.g., 700mA). If it's drifting more than 5%, consider the driver suspect.
"Why does this matter? Because an LED driver that fails open can leave you with a dark fixture — but a driver that fails at partial current can cause flickering, overheating, and even fire in enclosed fixtures. The multimeter test is your first line of defense."
The Real Lesson
After 5 years of reviewing quality complaints, I've come to believe that the 'best' vendor is not the one with the lowest defect rate, but the one that listens when the defect is found — even by a contractor with an $87 order and a $20 Harbor Freight multimeter. That customer still orders from us. Last week he submitted a $12,000 spec for the rest of the inn.
Small doesn't mean unimportant. It means potential. And sometimes, a dead driver is the best quality audit you'll ever get.