Why I Rejected an 'Almost Perfect' Coherent Verdi Laser

The number on my screen was 1.14. The pass/fail limit was 1.10. The test room hummed around me—cooling fans, the click of a relay from the next bench, someone's phone buzzing somewhere down the line. I'd been staring at the beam profile readout for the Coherent Verdi G for a full minute before I made any notes.

Not because the laser was broken. It wasn't. This Verdi G had produced a clean 532 nm green beam for eight straight hours. Power drift was negligible. Wavelength was dead-on. Thermal behavior was textbook. By any ordinary definition, it was a good laser.

But the M² value—that's the beam quality factor, a measure of how close the beam is to the perfect Gaussian ideal—was 1.14. Our internal threshold is 1.10. The difference is about 3.6 percent in focused spot area. Sounds like nothing, right? In most applications, it is nothing. You'd never see it on a finished part.

I've been a quality compliance manager at Coherent for four years. I review every system before it reaches customers—roughly 200 units a year. I've rejected about 7% of first deliveries in 2024 due to beam quality or alignment issues that would've compounded into customer-visible problems somewhere down the line. This was going to be one of them.

The Customer Behind the Serial Number

What makes a rejection harder is knowing who's waiting on the other end. This Verdi G was headed to a small manufacturer in Ohio that started as a one-person laser engraving business five years ago. The owner still talks about the day he called Coherent to ask what the best laser engraver for cutting boards was for under $9,000. An applications engineer talked him through his actual use case—custom cutting boards with intricate monogram work—and steered him away from the budget import he'd been eyeing online.

That decision set the tone for his whole trajectory. His cutting board line took off. He added a second CO2 laser. Then local manufacturers started calling about contract work, and he expanded into laser welding services using a Coherent fiber laser. Last year, he entered the OEM hardware space, building precision subsystems for a university research client. The Verdi G was going to be the pump source in one of those systems.

I visited his shop once. He had a whiteboard in the back with a handwritten motto: "The output is the brand." He told me it came from an early failure, when he delivered a batch of engraving that was "good enough" and lost the client to someone who charged more but delivered cleaner edges. He never forgot that.

The 54-Point Laser Check

The pre-shipment verification we run—what many customers call the Coherent laser check—is a 54-point protocol that takes about a full day. Warm-up curves, power stability over hours, beam pointing drift, wavelength accuracy, mode quality. We also take a thermal image of the resonator under load. Half of those readings never appear on a spec sheet. We do them anyway, because the failures we see in the field always leave fingerprints in the lab.

That day, the Verdi G passed 53 of 54 points. The one failure was the M² measurement.

I spent most of an hour going back and forth with myself before writing the rejection. Here's the honest math: a 3.6 percent larger focused spot size is invisible in a finished part. In a cutting board engraving, you'd need a magnifier to see edge softness. Even under the microscope used for weld inspection, it wouldn't flag. The practical consequence was close to zero.

But that's not the right way to frame it.

The right way: this customer's clients—the wedding-gift buyers, the medical device engineers, the university researchers—don't judge a product by whether it passed "within tolerance." They judge it by a single question: is it right? And "right" leaves no room for "3.6 percent off, but nobody will notice." Once a laser carries your brand into a customer's operation, you're making a promise about every part that comes off that machine. A laser that starts its life at the edge of its own spec is a laser that will drift over the edge when it gets warm, or ages, or gets a lens installed slightly off-axis.

I rejected the unit.

The Call I Didn't Want to Make

I won't pretend it was a pleasant conversation. The customer's production manager had a clear integration timeline. "Is this really necessary?" he asked. "The old system ran for thousands of hours without anyone checking the beam profile."

He wasn't wrong. But I've seen the alternative up close.

In 2022, I approved a pulsed laser that was borderline on pulse-width spec. Just barely within range—engineering had signed off. Eight months later, it came back to haunt us: the customer's output was showing weak weld seams on a precision enclosure job, and the fault traced back to a marginal measurement we'd let slide. That quality issue cost them a $22,000 redo and delayed their launch by a month. We didn't lose the account, but we came close. Looking back, I should have rejected that unit too. At the time, the shipping pressure made "within spec" feel like a reasonable call.

So I told this customer: "You'll probably never see a difference in your parts. But we're not in the business of shipping lasers that are probably fine."

Six Months Later

The rebuilt Verdi shipped two weeks late, with an M² of 1.06—better than the 1.10 spec, which is what we aim for. I didn't think much about it until I visited the customer's facility this past January.

The owner walked me through the shop. The cutting board line was running cherry and walnut stock for a pre-order that had sold out in four hours. The laser welding services station was full of fixtures for a local automation company. And in the back corner, the new OEM subsystem was humming—the Verdi G pumping a Ti:Sapphire cavity with a beam profile so clean it made me smile.

I asked the question I ask every customer: "Any quality issues since installation?"

"Zero," he said. Then he said something I won't forget: "I know you took a hit to send that laser back. Cost you time, cost your team a rework cycle. But the thing I've learned in five years of this business is that your reputation is just the accumulation of what comes off your machines looking right. One bad batch changes how people see you, whether they say it or not. You protected that for me. That's why I keep buying from you."

What the Rejection Taught Me

I've been turning that conversation over in my head for months. It's tempting to think quality control is about measurements, standards, and tolerances—and to a point, it is. But the real purpose of a coherent laser check isn't to catch broken lasers. Very few lasers come off our line actually broken. The purpose is to catch the things that are right on the edge.

Think of it the way the printing industry thinks about color: brand-critical colors are held to Delta E < 2 on the Pantone scale, because that's the threshold where a trained eye starts to notice a mismatch. The average buyer won't see the difference between Delta E 1.5 and Delta E 3.5 on one box. But across a print run of ten thousand boxes, the inconsistency becomes measurable. The same logic applies to the 300 DPI standard in commercial print—290 DPI looks identical on a single sheet, but on a long run the sharpness difference compounds. Laser quality works exactly the same way. M², power stability, pointing drift: these are your Delta E. The units we reject are rarely catastrophic. They're the ones that would slowly, invisibly degrade a customer's output until one day they're reworking parts and wondering why.

A Few Honest Caveats

My experience is based on reviewing about 200 Coherent systems a year, across the industrial and scientific range. I can't speak for every brand or every budget tier. If you're a small shop running a desktop engraver and just starting out, you don't have the option to demand M² 1.1 on a verification report, and the calculus is different. This approach worked for us because we're a laser manufacturer with the resources to hold a hard line.

But whether you're running a laser engraving business, offering laser welding services, or buying your first Coherent laser for a cutting board start-up, the principle holds: the quality of your output is the quality of your brand. Check your focus. Keep your optics clean. Measure your power at the work surface. Calibrate more often than you think you need to. And if a supplier tells you something is "within spec," ask a simple follow-up: how close to the edge is it?

Because the edge is where reputations get lost.