Thursday afternoon. Our packaging line had been running clean for three weeks. Then, at 2:47 PM, station 4 just stopped.
'The linear actuator's dead,' the maintenance lead said. 'Motor's still spinning, but the carriage isn't moving.'
If I remember correctly, that was the exact moment I started thinking about bearings. Because in my experience, when a linear actuator fails, the motor is rarely the real problem.
Diagnosing a failure that wasn't what it looked like
Our first instinct was the motor controller. We swapped the drive card, and nothing changed. Then we checked the belt—fine. We tested the guide shaft for binding—it turned freely by hand. That's when I asked to pull the actuator housing apart.
It's tempting to think a linear actuator is a simple device: motor, screw, nut, push. But the ball screw bearing and the linear ball bearing inside are the parts that actually carry the load. Our diagnosis made sense: the carriage was dead despite the motor spinning because the nut had lost its interface with the screw. The linear ball bearing cage had collapsed, and the balls had jammed against each other. Debris had scored the shaft. A textbook failure, except the textbook usually skips the part about poor maintenance history.
We checked the service log. Six months earlier, a maintenance tech had replaced the original koyo bearings cartridge with a cheaper 'equivalent' unit. The shaft fit, but the dynamic load rating was lower than what our cycle needed. On paper it looked fine. In service, it lasted four months.
That's the classic outsider blind spot: buyers focus on thrust rating and stroke length, and completely miss the tolerance class and load rating of the bearing inside. The question everyone asks is 'what actuator do I have?' The question they should ask is 'which bearing is in it, and is that bearing rated for the actual cycle rate?'
(I'm not an engineer, but after four years of quality reviews, I've documented enough bearing-related downtime to know where to look first.)
The deadline—and the 'probably fine' temptations
Once we had the diagnosis, the clock became the problem. The customer's final audit was Monday. Our regular distributor said a standard Koyo cartridge could take three to five business days. Monday was two days away.
The procurement manager suggested a reconditioned actuator from a local shop. It bench-tested fine. It was half the price. We could have it installed by Saturday afternoon. I had mixed feelings.
Part of me wanted the cheap, fast fix. Another part remembered a similar situation in 2023 when a 'compatible' bearing failed within nine days and cost us a $22,000 rework plus a week of downtime. I also knew we had a manufacturer's audit coming, and the auditor would likely check the replacement part numbers.
We asked the local shop for the bearing tolerance class. They couldn't confirm anything beyond 'it's the same size.' That word—same—was a red flag. According to ISO 492 (Rolling bearings — Radial bearings — Geometrical product specifications and tolerance values), the tolerance class controls things like radial runout and width variation. A bearing that looks identical can have significantly different running accuracy. For a precision screw, that difference matters.
So we made the call: order the proper Koyo parts and pay for overnight delivery. The freight was close to $400. I won't pretend I was happy about the number. But the extra cost bought certainty—a confirmed shipping number, a confirmed bearing spec, and no room for 'we think it'll work.'
I'm not saying rush service is always worth it. But when the cost of missing the deadline is a contract audit failure and a delayed launch, paying for certainty is the cheaper option. That's a lesson we've internalized the hard way.
The replacement and the Koyo logo check
The parts arrived Saturday evening. When the box opened, I looked for the koyo bearings logo on the packaging and on the inner seal. Not because I doubted the distributor, but because we'd had one counterfeit scare in the past. (Note to self: make that photo documentation part of the standard procedure.) The logo was clean, the part numbers matched the catalog PDF, and the raceway finish looked correct.
The technician replaced the collapsed linear ball bearing and the damaged ball screw bearing. Reassembly took about two hours. On the first power-up, the carriage moved with a smooth, even motion. Vibration readings stayed well below the alarm threshold. The line was running again by Sunday morning.
So glad we paid for that overnight freight. Almost went with the reconditioned unit to save a few hundred dollars, which would have meant risking a much bigger loss.
What I'd tell another maintenance team
If you're asking 'what happens when a linear actuator fails,' the short answer is: downtime, a confusing diagnosis, and—more often than not—a bearing that was overlooked during the last service.
Here's the part I wish someone had told me earlier:
- Don't condemn the whole actuator until you've inspected the bearings. A worn linear ball bearing can make a perfectly good motor and screw look completely dead.
- Specify the bearing tolerance class. For a ball screw bearing, the difference between ABEC 1 and ABEC 3—roughly ISO P0 and P6—affects how the load distributes across the balls and, eventually, how long the actuator stays quiet.
- When you're under a deadline, verify lead times in writing. If a supplier can't confirm a ship date, treat that as a risk, not an estimate.
- Keep one known-good spare in stock. The carrying cost is small compared to a weekend of emergency freight and second-guessing.
We now keep a spare Koyo cartridge in the cabinet and photograph the bearing markings during every service. It's not glamorous. But after you've watched a production line stop for 48 hours because of a $50 bearing, the habit sticks.
The Koyo bearings catalog PDF is saved on our maintenance tablet now, and our parts list has the Koyo part numbers next to every actuator. It's a small change. It has made the next audit conversation a lot easier.
