2026-11-10 - Gates engineering note

The $380 BLDC Motor Mistake That Nearly Cost Us a Client Demo

A procurement story from a 45-person engineering firm: what happened when I sourced a cheaper BLDC motor by unit price, and what it taught me about maxon-motor datasheets, 28byj-48 stepper motors, drive chain assemblies, and how ball bearings are made.

The spec sheet that landed on my desk on a Tuesday in March 2024

I'm the office administrator for a 45-person engineering services firm. I handle purchasing across roughly 12 vendors and about $1.1 million in annual spend. I report to both operations and finance, which means every order I place gets reviewed twice—once for need, once for cost. Most days, that's fine. Some days, it's the reason I make decisions I end up regretting.

On a Tuesday in March 2024, our lead engineer, Priya, dropped a spec sheet on my desk for a new material handling project. The list was longer than usual: maxon BLDC motors for the main actuators, four 28byj-48 stepper motors for a test jig, drive chain assemblies, ball bearings, and one special order—a maxon air s motor e-mtb for the client's demo bike.

"Can you source these by end of month?" she asked. "We're on a tight schedule, and the client demo is in six weeks."

I said yes. That was my first mistake.

What I did wrong (and why it felt right at the time)

Here's the thing about procurement at a company our size: the metric I was measured on was unit cost savings. My quarterly review literally had a line item for "cost avoidance." So when I started sourcing, I did what I'd always done—went looking for the cheapest option that looked close enough.

The 28byj-48 stepper motor was easy. Those are everywhere, and honestly, they're cheap everywhere. I found a seller offering a 20-pack for about $18 less than my regular source, with a two-day shipping guarantee. Ordered without thinking twice.

The drive chain components were also straightforward. I found a supplier who could match the specs at around 12% below my usual vendor, as long as I ordered the batch upfront. Ordered.

Then I got to the maxon BLDC motor line item and hit a wall. The published maxon-motor price per unit was higher than I expected. So I did what any cost-conscious buyer would do—I found a Chinese manufacturer selling "equivalent" BLDC motors at roughly $380 less per unit. Same voltage, similar torque rating, comparable frame size. The spreadsheet said I'd save about $1,520 across the four units we needed.

I ordered one sample unit to test, plus the full order of four. That was my second mistake.

Here's the rookie error, plain and simple: I assumed "BLDC motor" meant the same thing to every vendor. In my first year doing this, that assumption probably cost us a couple thousand dollars in redos. In this case, it nearly cost us a client.

The sample arrived with no datasheet. Not "a thin datasheet"—no datasheet. When I asked for torque curves, efficiency maps, and thermal specifications, the supplier sent back a single-page spec sheet with a torque rating and no testing methodology behind it.

Priya looked at it for about thirty seconds. "We can't integrate this into the actuator design without knowing the continuous torque at different ambient temperatures. This is a guess, not a spec."

We didn't have a formal vendor qualification checklist at that point—not really. We had a spreadsheet of past vendors and a rough sense of who'd burned us before. The third time a new supplier sent us something that didn't match its promised specs, I finally built a proper checklist. Should have done it after the first time.

The ball bearings that made me learn something I never expected to need

While all of that was happening, the drive chain assemblies arrived. Bench testing started on a Thursday. By the following Tuesday—about 40 hours of testing—one of the ball bearings in the assembly showed visible wear. Not a little wear. Pitting on the raceway.

I had no idea what to look for. So I spent about two hours that evening Googling "how ball bearing made"—the actual manufacturing process, not just the marketing copy. It turns out the difference between a cheap bearing and a quality one often comes down to the grade of steel and how carefully it's heat-treated. The cheap ones I'd bought were almost certainly not GCr15 bearing steel, or if they were, the heat treatment was rushed. You can't see that from a spec sheet. You can only see it when the raceway starts pitting at 40 hours instead of 4,000.

The supplier's response to my complaint was, and I quote, "The bearing working." They technically weren't wrong. It was working. It just wouldn't be working after another 200 hours, which is roughly when our client's system would go live.

The moment the spreadsheet stopped being the answer

Here's where the data and my gut disagreed.

The spreadsheet said the cheap BLDC motor saved us $1,520 across four units, plus the drive chain batch saved another $220, plus the 28byj-48 "deal" saved $18. Total "cost avoidance" on paper: about $1,758. That's a number I could have put in my quarterly review and gotten a nice nod from finance.

My gut said something was off. Not because I'm some expert on motors—I'm not. But the same pattern had played out before. Cheap item looks fine, fails under real load, we scramble, the client notices.

I ordered one maxon BLDC motor as a comparison unit. Full price. No negotiation. When it arrived, it came with a printed datasheet that listed continuous torque at 25°C, 40°C, and 60°C ambient, efficiency curves at four operating points, and thermal resistance values for the housing. There was even a note about Swiss manufacturing tolerances that I didn't fully understand, but Priya did.

We ran both motors through the same continuous load test. The cheap motor overheated and cut out at about 60% of its rated continuous load. The maxon unit held steady through the full test cycle, and its housing temperature stayed within about 8°C of what the datasheet predicted.

That comparison cost me $600 for the maxon unit. It saved the project.

Never expected the specification gap to be that wide. Turns out "equivalent" and "compatible" are very different words when you're talking about a motor that runs eight hours a day inside a client's production line. And maxon publishes its continuous torque ratings at defined ambient temperatures—that's become my baseline for any motor vendor now. If they can't tell you what the motor does at 40°C, they don't know.

What I do differently now

We ended up ordering the full set of maxon BLDC motors. The drive chain assemblies were rebuilt with bearings I ordered from a specialty supplier that actually publishes the steel grade and heat treatment process—which, after my evening reading session, I now know is the minimum bar for anything with a duty cycle above a few hundred hours.

The 28byj-48 steppers? Fine. Those are for a test jig that runs a few hours a week. The cheap ones are probably fine for that specific use, and they've held up. Not everything needs to be premium, and I don't think anyone should pretend otherwise.

The maxon air s motor e-mtb for the client demo arrived with full documentation, on schedule, and the client signed off on the demo within a week. I'm not 100% sure that would have happened if I'd tried to substitute that one for something cheaper. Take that with a grain of salt—I only have one data point on the e-MTB side. But it did what the datasheet said it would do.

But the broader lesson stuck: unit price is one number. Total cost is everything else. The $380 I "saved" per BLDC motor would have cost us the project, and probably the client relationship. That's a calculation I should have done upfront, and one I now do on every order that goes into a client-facing system.

Before I place any order like that now, I check three things: Can I get a real datasheet with test methodology? Has the specification been validated at the operating conditions we actually expect? And what does the failure mode look like—does it degrade gracefully, or fail all at once?

If a supplier can't answer those questions before I've signed a PO, they're not the right supplier, no matter what the quote says. That's not a lesson I learned from a textbook. It's one the pitting on that ball bearing taught me.


Ask about this topic

Send belt drive requirements

Share pulley diameter, center distance, rpm, ambient temperature and duty cycle. A specification engineer will return a compact selection path.

By submitting this form you agree to the Privacy Policy.