As office administrator for a 45-person equipment integrator, I manage around $150,000 in annual purchasing across 11 vendors. Since 2020, I have ordered everything from bearing blocks to complete linear motion systems. The question that comes up most from our engineers: should we buy a fully engineered thomson-linear actuator, or build it ourselves from a motor, a linear shaft, and a screw?
This isn't a lazy 'it depends' article. I have a framework. We compare the two approaches across four dimensions: total cost of ownership, reliability, technical support, and integration risk. If you're trying to decide between a commercial actuator and a custom build, this should make the real trade-offs visible.
The comparison framework
Option A is a complete thomson linear actuator with an integrated motor, feedback, and limit switches. Option B is an in-house assembly using thomson linear shafting, a ball screw, and a motor—often a century ac motor for constant-speed moves, or an mg995 servo motor for prototype positioning.
Why compare these two? In 2024, during our vendor consolidation project, I pulled 14 months of purchase orders. We had spent about $61,000 on engineered actuators and $38,000 on individual motion components. The components looked cheaper per line item. But the engineering time to integrate them never appeared on a PO. That was the first clue that we were comparing the wrong numbers.
Dimension 1: Total cost of ownership
At face value, the DIY route wins on price. A bare mg995 servo motor might cost $10 to $15 online. A used century ac motor from a surplus catalog can be $75 to $200. Add a Thomson linear shaft, bearings, a screw, and an aluminum frame, and you might be at $300 to $700 for one axis. A complete thomson linear actuator with an integrated motor and feedback? Based on quotes from our distributors in January 2025, figure roughly $900 to $2,800 depending on stroke, load, and IP rating. That's real money.
Here's what the simple comparison misses. The $10 mg995 needs a driver, a controller, wiring, connectors, and a mounting bracket. Then someone has to figure out why it jitters at low speed. The cheap century ac motor needs a VFD if you want variable speed, and it has no position feedback at all. By the time we added the supporting parts, our first DIY axis cost about $1,100 in components and 30 hours of engineering time. At a loaded rate of $85 per hour, that's another $2,550. The Thomson actuator cost more on the PO, but it arrived with a matched motor and wiring diagrams. (Should mention: we did wait extra for a custom cable, but the design lead time was shorter than our internal debugging loop.)
I'm not saying DIY never makes sense. But if you're comparing unit prices, you're comparing the wrong number. The 'just buy the parts' advice ignores integration and debugging costs.
Dimension 2: Reliability and duty cycle
What most people don't realize is that reliability isn't a single test result. It's a combination of material, tolerances, and how components are matched. If you've ever looked up how ball bearing made, you know the process—wire forming, race grinding, heat treatment, precision assembly. That's why a bearing from a real linear motion company and a no-name bearing can look identical and perform completely differently.
The same is true for shafting. The thomson linear shaft is induction-hardened, ground, and straight to a documented tolerance. In one DIY project, we initially used generic chrome shaft. It looked fine. The linear bearing bound after six weeks. We switched to Thomson shafting and the problem disappeared. I might be misremembering the exact spec from the datasheet, but I believe the straightness tolerance was about 0.001 inches per foot. Whatever the number, it mattered.
The motor gap is even bigger. An mg995 servo motor is a hobby component—it has plastic gears and a small motor meant for occasional movement. We used one in a low-load positioning prototype, and after four months the gears stripped. In hindsight, we should have known better. A century ac motor, on the other hand, is an industrial workhorse for continuous duty, but it isn't a positioning device. It runs, or it runs at whatever speed a VFD tells it. If the application needs to hold position under load, it's the wrong tool.
Our pre-engineered thomson linear actuator has been running on a bagging line for three years. It gets roughly 20,000 cycles a year, and the only maintenance has been checking fasteners and cleaning the shaft. That's anecdotal, but it matches the lifecycle data Thomson publishes. According to Thomson's product documentation (thomsonlinear.com, accessed January 2025), the actuator's rated life at our load was well above our requirement.
Dimension 3: Technical support and honest boundaries
This is the part that changed how I buy. In 2022, I called our distributor about replacing a failed motor on a conveyor. The failed unit was a century ac motor, and I assumed the solution was another motor. The Thomson applications engineer asked what the machine was doing. After a few questions, he said, 'The motor isn't your problem. If you keep the existing motor, replace the linear shaft and the bearings, and check the coupling alignment, you'll likely be fine.'
To be fair, they could have sold me a motor. But they didn't. That 'this isn't our strength—here's what to check' moment earned my trust. Later, when we needed a multi-position actuator, they told us their screw-driven actuator was the right fit. For a high-speed light-load move, they suggested a different product because their own option would have been overkill. No one in industrial sales says that unless they're confident in their actual niche.
I'd rather work with a specialist who knows their limits than a generalist who overpromises. That's true for vendors, and it's true for our own engineering choices.
Dimension 4: Integration and the one-stop trap
Here's something vendors won't tell you: 'universal compatibility' is a red flag. An mg995 servo motor uses PWM signals and small connectors. A century ac motor uses line power and needs contactors or a VFD. A thomson linear actuator can be ordered with an integrated servo or stepper motor and feedback, and its mounting pattern is documented to the millimeter. Those are not interchangeable.
When we built the DIY axis, we spent two weeks making the motor and the screw shaft align. The motor bracket, the coupling, the shaft support, the end bearing—each came from a different supplier, and each had slightly different tolerances. A pre-engineered actuator eliminates most of that because the motor, gearbox, screw, and carriage are designed as one system.
That doesn't mean you should never integrate components yourself. It means the 'piece it together from a catalog' approach only works if you have the engineering time and the ability to debug. We do it for prototypes. For production machines, we buy the complete actuator.
Which should you choose?
Here is my honest rule of thumb.
Choose the complete thomson linear actuator if your application runs many cycles a day, needs position or speed control, or if your team doesn't have a controls engineer who enjoys chasing EMI from a hobby servo. This is also the route when you need documentation—certificates, IP ratings, installation manuals. For a machine that ships to a customer, that paperwork matters more than the part cost.
Consider the DIY route if you're building a proof of concept, the motion is simple and constant-speed, and you have in-house experience. For a two-position transfer, a century ac motor with a limit switch and a Thomson linear shaft can be a perfectly good solution. For a low-cost robot demo with no industrial interface requirement, an mg995 servo motor might be enough. Just budget for the first prototype to fail.
And if you aren't sure which category you're in, ask the vendor. A good one will tell you which of their products is overkill and which is insufficient. That kind of conversation saves everyone time.
Prices as of January 2025; verify current rates with your distributor.