Most 'Failed' Danfoss VFDs Aren't Dead: What Alarm 60 and A14 Really Mean

I'm going to say something that might not sit well with some of my fellow drive specialists: most emergency VFD callouts never should have been emergencies in the first place.

In my role coordinating emergency repairs for manufacturing clients, I've fielded north of 200 drive-failure calls in the past year. Maybe 180, I'd have to check the log. And the share that actually needed a replacement drive? Probably about a third, give or take. The rest were wiring faults, interlock loops, and ground leaks that had been building up quietly for weeks.

The industry has moved on, and a lot of troubleshooting habits haven't caught up. Modern VFDs like the Danfoss FC series are basically self-diagnosing. They spit out a code that tells you exactly what's wrong. The problem is that too many of us treat every fault code like it means "replace me." It usually doesn't.

Danfoss VFD Alarm 60 — External Interlock: The Drive Is Doing Its Job

Danfoss alarm 60, the external interlock fault, is one I see constantly. Per Danfoss's own documentation (the FC Series operating guides on danfoss.com), alarm 60 means the drive has lost permission to run — the interlock input is open, so the drive stops and reports it. That's not a failure. That's obedience.

The interlock loop is part of your safety circuit: emergency stops, panel door switches, safety relays, sometimes a PLC signal. When anything in that chain opens, the drive declares A60 and won't restart. The system is working exactly as designed.

So what actually causes A60 in the field? In my experience — and this is mostly food & beverage and packaging plants, I should note — the usual culprits are:

• An e-stop button that got pressed and didn't re-arm properly
• A panel door interlock switch that's out of position or broken
• A safety relay that dropped out, or a loose wire on the interlock terminals
• A PLC output that isn't energizing when the sequence calls for it

The worst A60 call I can remember happened in March 2024, 36 hours before a client's production audit. Their entire packaging line stopped at 11 PM. The maintenance manager was convinced the drive was dead and wanted a replacement by morning. Normal turnaround on a drive evaluation is two to three days, so I stayed on the phone and walked him through the interlock check.

"Is any e-stop pushed in?"

Silence. Then: "Uh... the one on the panel by the wash bay. It's in."

A cleaning crew had hit the mushroom button, and nobody re-armed it. The emergency was over with a twist of the hand. But the client paid a $400 callout fee for what a ten-second visual check would have solved.

There's a second gotcha worth knowing: on many Danfoss drives, if the interlock is set to latched, clearing the alarm isn't enough. You have to restore the interlock signal and then reset the trip before the drive will run. I've watched experienced electricians swap an entire drive because that detail slipped past them.

One more tip: if A60 reappears immediately after every reset, the interlock input itself is open — the circuit is broken, not the drive. Check continuity of that loop before you even think about the drive. And check the fault history — Danfoss drives log recent alarms, and the sequence leading up to the trip often tells you more than the code alone.

Danfoss VFD Earth Fault A14: The 15-Minute Test Before You Replace

Alarm A14, the earth fault alarm, is a different story. And this is where my "the industry has changed" argument gets concrete.

A motor that ran fine on a 1990s VFD can trip A14 on a brand-new Danfoss drive. Not because the motor suddenly got worse — because the new drive measures better. Faster sampling, more accurate current sensing, more sensitive ground-fault detection. The new drive is more honest than the old one.

What causes A14 in practice?

• Motor cable insulation damage. Cables dragged over sharp edges, soaked in oil for years, chewed by rodents — all of that shows up eventually.
• Moisture in the motor junction box. That's the big one in washdown environments. Condensation plus a motor that cycles warm and cool adds up to a low-grade ground leak.
• An old or overheated motor with degraded winding insulation.
• A genuinely faulty drive. That's the least common cause I've come across, honestly.

Here's the no-cost, 15-minute diagnostic: disconnect the motor cable from the drive and run the drive with no load. If A14 clears, the problem is in the cable or the motor. If it stays, you're likely looking at a real drive fault. That's a test that separates a $50 cable repair from a $2,000 drive replacement. I've seen too many facilities skip it and order the expensive part first.

Another thing that surprises people: newly wired or reworked systems trip A14 more often than old ones. You can have a subtle ground loop — two different ground potentials between the drive chassis and the motor frame — and on a sensitive modern drive, that's enough to trigger the alarm. Properly bonding the motor cable screen at both ends usually clears it up.

Honestly, I'm not sure why some plants get chronic A14 trips while similar facilities with the same equipment run for years without a single one. My best guess is cable routing discipline and how carefully junction boxes get sealed (corrosion in terminals is a close second). But I've also been to spotless plants that can't shake earth faults. If someone has a better theory, I'd genuinely like to hear it.

Same Pattern, Different Parts: Fanuc Servo Motors and LM8LUU Linear Bearings

This "diagnose before you replace" thing isn't unique to variable frequency drives. It runs through the whole automation industry, and it's become more obvious as components got smarter.

Take Fanuc servo motors. When a servo amplifier throws an overload alarm, the code points at the axis. But in a surprisingly large share of the calls I get pulled into, the real problem is mechanical — a stuck linear guide, a misaligned ball screw, a bearing that's dragging under load. The servo motor is doing what a servo motor does: working harder, trying to follow the command, and complaining loudly about it. Replacing a $3,000 motor doesn't fix the stuck bearing. It just gives the machine a new motor with the same old problem.

And this is where servo motor control matters. A properly tuned axis will tell you when it's fighting the mechanics. The current draw, the following error, the alarm code — all of that is the system saying something downstream is wrong. Read it before you throw parts at it.

Then there's the search query that sums up the whole problem: "what size is lm8luu linear bearing."

The answer: LM8LUU is a metric linear bearing with an 8mm inner diameter, 15mm outer diameter, and 45mm body length. The "UU" means double-sealed; the "L" designation means it's the flanged long version, which provides more load support in linear motion assemblies. A 30-second lookup with that part number saves you from ordering the 10mm version that doesn't fit and turning a 20-minute swap into a two-day wait.

That's the evolution in one sentence: the industry moved from "I have this memorized" to "I know how to check this" — and the second skill is harder for old-school engineers than they'd like to admit. What was best practice in 2010 is not automatically best practice in 2025.

But Some Drives Do Fail. I've Replaced Plenty.

Let me address the obvious objection before you make it: "Are you saying drives never die?"

No. Absolutely not. Semiconductors fail. Power surges kill IGBTs. Capacitors have a finite lifespan. Last quarter alone, we processed 47 rush orders for replacement drives and automation components, and a good number of those were real failures — the kind where the drive smells like burnt electronics and the DC bus reads zero. In those cases, replace it, and fast. I'll be the first to defend a same-day swap when the drive is genuinely dead.

But here's the edge case that keeps me up at night: if the fault is external and you replace the drive anyway, you haven't fixed anything. The broken interlock circuit is still broken. The wet junction box is still wet. The damaged motor cable is still leaking current. And the new drive — more sensitive, more honest — will catch the fault even faster and trip again.

I made this exact mistake once, early in my career. I assumed that because the replacement was the same model, I could swap it and move on. Didn't verify the motor circuit. Turned out the old drive had been masking a damaged cable — and the new unit detected it on the first energization. Watching a brand-new drive trip A14 in front of a client is a specific kind of professional embarrassment that stays with you.

That's why our company now operates under a simple rule:

"No replacement drive gets installed until the interlock circuit and the motor circuit are both verified."

That rule came from a 2023 incident where we paid $800 in rush fees and still lost a day of production because we didn't follow it. Never again.

And look, the supply chain part has gotten better, too. Five years ago, a same-day drive swap was a logistical miracle. These days, with a decent distributor, it's almost routine. But that speed shouldn't replace the diagnostic step — it should come after it. Ordering a drive is easy. Confirming it will actually solve the problem is the part that saves you from doing this twice in one week.

The Bottom Line

The fundamentals of motor control haven't changed — a VFD has to deliver voltage and frequency to a motor, and a bad circuit is always bad. But the execution has transformed. Drives diagnose themselves with impressive precision, if you actually read what they're telling you.

So before you call for an emergency replacement:

1. Check the interlock circuit. Any e-stop pushed in? Door switch seated? PLC output present?
2. Disconnect the motor and run the drive solo. Does A14 clear?
3. Look up the part number. LM8LUU is 8mm, not 10mm. The Fanuc alarm code has a meaning. Read the manual.

Diagnose first, replace when you have to. Most "failed" Danfoss VFDs — the A60s, the A14s, the whole category — aren't dead. They're waiting for someone to read the code and check what's actually wrong.