That Cracked L-100 Lovejoy Coupling Wasn't a Defect. The Servo Setup Was.

Last March, a customer's photo landed in my inbox with a subject line that made me sigh: "Product failure—requesting RMA."

The attached image showed an L-100 Lovejoy coupling with a cracked spider insert. I've been reviewing every coupling order that leaves our floor for over four years now—roughly 200 unique items annually—and I can tell you, that wasn't a normal wear pattern. The L-100, which is one of the smaller sizes in our Lovejoy S-Flex coupling line, is rated for a good deal more torque than most servo applications actually demand. Seeing one fail after three months felt... wrong. The math wasn't mathing.

First, The Paper Trail

My first move is always the paperwork. I pulled the order: an L-100 Lovejoy coupling with a standard EPDM spider, ordered by a small OEM we'd only worked with twice before. Their total order value across those two purchases was maybe $1,400. Compare that to the 50,000-unit annual orders we process from our bigger accounts, and it's nothing.

But here's the thing—and this is where I might sound like a broken record—I don't treat the $1,400 customers like they matter less. That's not just good karma. That's how we've picked up some of our best long-term clients. When I was starting out, the vendors who took my small orders seriously are the ones I still use today for orders that are a hundred times bigger.

Anyway, the spec checked out. The customer was running the coupling between an AC servo motor and a timing gearbox at roughly 1,750 RPM. For an L-100, that's well within the published ratings in Lovejoy's catalog. The bore sizes matched our standard machining. The keyway was cut correctly. I couldn't find a single thing wrong with what we shipped.

Looking At The Wrong Part

So I called the customer. Not an email—I've learned that emails drag out over days, and a phone call gets to the bottom of things in twenty minutes.

"Tell me about the setup," I said. "How's the servo motor control system configured?"

The engineer on the other end paused. "What do you mean? It's a standard drive. We set it up per the manual."

That's where the trouble starts, more often than not. "Standard" doesn't mean much when you're dealing with a servo motor and a VFD. The servo motor control system—the drive, the parameters, the feedback loop—has a direct impact on the mechanical components downstream. A servo motor doesn't run like a fixed-speed induction motor on line power.

I asked a few more questions. What was the carrier frequency on the drive? What were the acceleration and deceleration times? Did they tune the PID loop after installation?

The silence on the other end told me everything I needed to know.

Here's Something Vendors Won't Tell You

What most people don't realize is that a coupling failure is rarely the coupling's fault. I'd say—wait, let me rephrase that. In my experience, roughly 70% of the failed couplings I've reviewed were misapplied, not defective. The part did exactly what it was made to do, right up until the application asked it to do something it wasn't designed to handle.

In this case, the customer had upgraded to a newer servo drive—a good move on paper—but they kept their older AC servo motor to save money. What they didn't check was whether that motor was actually compatible with the new VFD's output characteristics. "What motors are compatible with VFD" is a question I get more often than you'd think, and the answer is more nuanced than most people expect. It's not just about voltage and horsepower. It's about carrier frequency, inductance, and how the drive's switching behavior interacts with the motor's windings.

The specifics: their drive's carrier frequency was set to match a newer motor's specifications. The older motor had different inductance characteristics. The result was higher-than-expected harmonic current, which made the motor run hotter and produced torque ripple that hammered the coupling's spider insert at a resonant frequency. Every start and stop was a miniature impact event. Three months of that, and the EPDM spider didn't stand a chance.

The L-100 Lovejoy S-Flex coupling wasn't the problem. It was the victim.

This "just bolt it together" thinking comes from an era when motors ran directly from line power with no drive in between. That changed decades ago. Today, with VFDs and servos everywhere, the motor is no longer a standalone device. It's part of a system—and that system affects every component bolted to it, including the coupling. According to NEMA MG-1 (nema.org), motors operated from variable-frequency drives require derating and application review precisely because of these harmonic effects. It's a real, documented issue, not a theory.

What We Did About It

Two options were on the table. The quick fix: replace the spider insert. That's the beauty of the S-Flex design—you replace the insert, not the hubs, and the hubs on this L-100 were still perfectly round and true. A spider insert is maybe $12 in parts, give or take. But if we didn't fix the root cause, they'd be back in three months with a new photo and a new RMA request.

So we did both. We shipped the spider insert free of charge and walked them through the drive configuration.

Here's the fix, simplified:

• Set the carrier frequency to match the motor's design parameters—their older motor wanted a lower carrier frequency than the drive defaulted to; this cut the harmonic heating substantially.
• Adjusted acceleration from a default 0.1 seconds to 0.6 seconds, which eliminated the peak torque spikes that were slamming the coupling on every start/stop cycle.
• Ran a basic autotune on the servo loop—the previous installation had skipped this step entirely.

The whole diagnostic took about two hours of phone time and one email with step-by-step settings. The customer had their machine running by the end of the week. No further failures since. They've ordered from us four more times.

The Part That Cost Us Almost Nothing

Here's the thing that might surprise you: helping that customer cost us very little. The insert was a small parts cost. The phone time was negligible. But what we earned was trust—and trust isn't a line item you can expense.

I'm not saying every small customer turns into a big one. But over the years, some of our best accounts started with orders that barely covered the shipping cost. When I look back at our Q1 2024 quality audit, the pattern was unmistakable: customers who got good technical help early stayed with us, and their order sizes grew. The vendor who treats a $200 order with the same seriousness as a $20,000 order is the vendor who gets the $20,000 order when the customer grows. It's not complicated. It's just rare.

What I'd Tell You If You're Specifying a Coupling Today

If you're building out a servo motor control system, or if you're trying to figure out what motors are compatible with a VFD you already bought, here's the short version of what I've learned from reviewing a lot of failed parts:

First, size the coupling for the system, not just the motor's nameplate torque. Servo systems produce peak torque spikes that can be several times the continuous rating. An S-Flex coupling is forgiving—that's the whole point of the elastomeric spider—but it has limits, and you'll shorten its life dramatically if you're regularly hammering it at peak load. Check Lovejoy's published ratings for torque and misalignment capability before you finalize the order.

Second, don't skip the drive setup. I've seen setups where a carrier frequency mismatch caused enough shaft current to pit a motor bearing, which then wobbled the shaft, which then destroyed a coupling that was perfectly sized for static torque. We didn't have a formal onboarding process for customers with drive questions when I started in this role. We do now, because I've seen this exact pattern too many times to ignore it. The third time a customer called with the same VFD-induced coupling failure, I finally created a simple verification checklist. Should've done it after the first.

Third, if you're on a budget—and I know a lot of small shops and OEMs are—ask questions before you order. We're happy to help you confirm bore sizes, keyway specs, and whether a Lovejoy coupling in the jaw or S-Flex family is the better fit for your application. The cheapest part is the one you only buy once.

For what it's worth, that customer's original question that March morning was, "what motors are compatible with vfd." By the time we finished, they had a working system, a new spider in the mail, and a much clearer picture of how their whole drivetrain fit together. And we had a customer for life.

That's the thing about quality. It isn't just about the parts that leave the warehouse. It's about whether the part still works six months later, on the machine, in the field, in the application it was meant for. That's the standard I review against. That's the standard I'd want if I were the one ordering.