If you're designing a linear motion machine for a living, you know the drill: first you look at the load, then you pick an actuator, and only later—usually on the way to the ordering screen—you wonder whether the motor is good enough.
Stop there. Pick the motor before the actuator. Not the other way around. And I mean that regardless of which linear actuator brand you're looking at.
I didn't always think that. My first few years specifying Thomson linear actuators were a running series of expensive lessons about what happens when you treat the motor as the last item on the BOM.
I manage motor and actuator specs for a small automation company. Since 2019, I've personally made (and documented) 14 significant motor-spec mistakes, totaling roughly $20,000 in wasted budget. Now I keep a checklist so the rest of the team doesn't repeat them. Bottom line: the motor is the identity of the axis, and the actuator is the muscle. Muscles don't work if the brain has no idea.
Mistake No. 1: Treating a Small Servo Motor Like a Generic Accessory
Early in 2022 (or was it late 2021? — I'd have to check the spreadsheet), I ordered 16 Thomson linear actuators for a pick-and-place frame. The customer's motion profile looked straightforward: move, pause, move. I specified a small servo motor for each axis because the word servo felt like a safe engineering answer.
Here's what you need to know: a small servo motor's rated torque is not the number that matters. The reflected inertia is. I assumed the load would be well within the motor's inertia ratio. Didn't verify. Turned out the moving mass was about five times the motor's comfortable rotor inertia. On the first dry-run cycle, the motor spent its time oscillating instead of moving.
The actuator was fine. The coupling was fine. The motor was fine. The combination was not. That's the part nobody puts on a datasheet.
We caught the issue before the final build, but the fix wasn't pretty: 16 smaller-bore actuators? No wait—the fix meant a larger gearbox and a different motor. Around $2,800 in reorder costs, plus a two-week delay. Actually, three weeks, because the new motors had a lead time.
What I mean is: don't choose a motor because it looks like it should work. Choose it by the load-to-rotor inertia ratio, torque at the required speed, and duty cycle. If someone quotes you a Thomson linear actuator before asking about inertia, they're selling, not engineering.
The Single Phase AC Motor Is Boring, and Boring Is Fine
The opposite mistake is also common—or rather, it becomes a mistake when pride decides. Engineers love a small servo motor because it looks sophisticated. But a lot of linear motion doesn't need sophisticated. It needs a motor that starts, runs, stops when told, and does it again tomorrow.
In one project, a cart lift had a fixed speed, fixed travel length, and no position servoing in the middle. The original spec called for a small servo motor and a controller. I asked the obvious question: what is the servo doing that a single phase AC motor can't do? The answer was nothing useful.
We swapped in a single phase AC motor with a proper gearbox and a torque-limiting coupling. The motion controller disappeared. The tuning debate disappeared. The machine did exactly what it was supposed to do—boringly, reliably, cheaply.
Saved about $360 per axis on that one. Maybe $400, give or take, once we cut the cable lengths and the programming time. The real saving: we didn't hire someone to tune a servo that was overkill from the start.
Now don't get me wrong: a small servo motor belongs wherever loads change, positions vary, or speed must be precise. But the single phase AC motor is a no-brainer for a large chunk of simple transfer and lift applications. Simple is not a dirty word.
What Stepper Motor Should You Use? Stop Asking That First
A lot of people search the web with some version of what stepper motor should I use for my linear actuator? It's a fair question, but it's missing the second half: should the motor be a stepper at all?
Steppers are a great middle ground—cheaper than servos, simpler to control, and perfectly good for indexing between points. I use them more than anyone expects. But the open-loop part is a red flag if you ever lose steps. Trust me on this one: a lost step on a vertical axis can be a safety incident, not just a quality issue.
If you need to hold position while idle, a stepper with a brake and a Thomson linear actuator is a classic combination. If you need high torque at high speed, or closed-loop confirmation, you should be looking at a small servo motor instead.
And if you're comparing what stepper motor options by frame size only, stop. The frame size tells you the bolt pattern. The torque-speed curve tells you whether the pinion will actually move under load. That curve, not the frame size, is the spec worth your time. (In other words: NEMA 17 and NEMA 23 are envelopes. They aren't performance ratings.)
Also—and this was a painful lesson—never assume the torque-speed curve on the website matches the voltage at the machine. We did once, on a 200-piece order, and every unit underperformed. We saved maybe $8 per motor on the initial quote. The budget motor choice looked smart until we saw the temperature. Net loss: roughly $1,100 in labor and expedited replacements.
What About Letting One Supplier Solve Everything?
Here's the usual pushback: why should I care which motor to pick? Let the linear motion supplier quote the whole system.
I'm not against packaged solutions. I've seen motion control components from one supplier work cleanly when the supplier knows the full application. Thomson, for example, has the portfolio to supply the linear actuator, the motor, the coupling, and the controls. That's a real advantage when the engineering is done honestly.
But we can supply everything is not the same as we sat down to verify the inertia, duty cycle, and speed profile. A packaged system is only as good as the weird variables you gave them. If you don't know your own machine's duty cycle, no supplier can guess it for you.
I've also got a rule about boundaries: I'd rather work with a specialist who knows their limits than a generalist who overpromises. I'm not a motor design engineer, so I can't speak to winding harmonics or cogging torque analysis. What I can tell you from an applications perspective is that the most credible suppliers are usually the ones willing to say this isn't our strength or that spec needs a different specialist. That honesty is a game-changer.
Thomson publishes a lot of the relevant numbers in their sizing tools and technical documents, and it's worth checking rather than trusting my memory. But even the best data doesn't remove the need to define the motion profile first.
Start With the Motor, Then Trust the Right Kind of Specialist
So here's my updated checklist, hard-won from too many bad calls:
- Define the motion profile before contacting anyone.
- Pick the motor based on inertia, torque at speed, and duty cycle.
- Then choose the linear actuator—maybe a Thomson, maybe not—that matches the motor's real output.
- Ask the supplier about their boundaries, not just their catalog.
I've written Thomson linear actuator in more POs than I can count—maybe 400, perhaps 380, I'd have to check. I like the product line. But I like it even more when the application is sized in the right order.
Bottom line: the motor comes first. The stepper, the small servo motor, or the single phase AC motor—each one is a different personality, and you have to know which one your machine needs before you buy the arm that carries it. Take it from someone who has the mistake log to prove it.