We Kept Blaming the Motors. The Real Problem Was How We Ordered Them.

The Order That Went Wrong (Again)

6:40 a.m., a Tuesday in March. The email from our maintenance lead was three words long: "motor's down again."

I'm the office administrator at a 70-person machining shop. I don't have an engineering degree. What I have is a purchase-order system, a shared inbox, and about 11 vendors spread across everything from coolant to cutting tools to fasteners to motors. Roughly $340k a year passes through my desk, and I report to both our operations manager and to finance.

That motor was the third one in fourteen months. Same sequence every time: breakdown, panic, rush order, part arrives, part doesn't fit or doesn't last, another email.

For most of that stretch I thought we had a motor problem. Cheap motors. Bad luck. Whatever the explanation was, it wasn't my problem — I bought what I was told to buy, at the price finance wanted to see.

Then I started keeping a spreadsheet. (I really should've done this two years earlier, but here we are.) The pattern was hard to miss: the failures clustered around the orders I'd placed on price alone, from a vendor I'd never spoken to, using a part number somebody handed me in a hallway.

The motors weren't the problem. How we were buying them was.

What's Actually Going On

It took me a while to see this, because the failure always looked like a product failure. Motor burns out, motor stalls, motor runs hot. You replace the motor. That's the loop. But the loop keeps repeating because nothing upstream changes.

1. We were ordering part numbers instead of applications

Here's what an order request from the floor usually looked like:

"Need a NEMA 17 stepper, 1.8 degree, 40 oz-in. Two of them."

So I'd order exactly that, and it would arrive, and it would run hot and stall under load. And everyone would shrug, because the spec had been met. On paper, we got what we asked for.

What I didn't understand until a distributor walked me through it: NEMA 17 is a faceplate dimension. It means the motor is 1.7 × 1.7 inches square where it mounts. That's it. It says nothing about stack length, current rating, holding torque, or what driver you're pairing it with. Two NEMA 17 stepper motors with the same 1.8° step angle — 200 full steps per revolution either way — can differ by three times in available torque. Neither one is wrong. One of them is wrong for your job.

Same story with 3-phase AC motors. "We need a 3-phase AC motor" isn't a specification, it's the beginning of a conversation. Frame size, horsepower, pole count, voltage, mounting configuration, enclosure type, insulation class, duty cycle. Pole count matters more than people expect: a 4-pole induction motor on 60 Hz supply has a synchronous speed of 1,800 rpm, and at full load it'll turn somewhere around 1,750 with normal slip. Order a 2-pole when the application needed 4-pole and you'll find out the expensive way. Nothing on the invoice will look wrong.

And then there's the question I got from a new hire last year, which I initially thought was a joke: what's a servo motor? It's a fair question, actually, and it took me an embarrassing amount of time to answer it properly. A servo motor is a motor with feedback — an encoder or resolver telling the drive exactly where the shaft is, so the drive can correct in real time. You don't size a servo on horsepower. You size it on continuous torque at a specific speed, and you keep peak torque as a separate number for acceleration events. Size it on peak and it'll be fine in January and heat-soaked by July.

2. Nobody owned the whole system

Motor. Drive. Gearbox. Load. Four different conversations with four different people, sometimes four different vendors. That's where the finger-pointing lives.

The piece I'd been completely blind to: if a motor runs off a variable frequency drive, the motor isn't seeing a clean sine wave. It's seeing a PWM output switching at a few kilohertz, and depending on cable length and drive settings, that can put voltage spikes at the motor terminals well above what the nameplate implies. Insulation systems have to be qualified for that environment. This isn't a marketing label someone invented — it's a test standard.

NEMA MG 1 Part 31 defines what an inverter-duty motor actually is, both thermally and electrically. IEC 60034-18-41 covers insulation systems for machines fed from voltage converters. These are qualification standards, not adjectives.

When I bought a standard-efficiency 3-phase AC motor for a drive-fed application because it was $70 cheaper, I wasn't saving $70. I was borrowing against a future failure with interest. It took two of those to make the connection.

3. Small orders got treated like small problems — and I was part of the problem

This one's on me.

For a long time I avoided calling distributors about a two-motor order. It felt like wasting their time. So I'd go to whoever had stock and didn't ask questions, order two units, and move on with my day. Efficient, right up until it wasn't.

What I eventually figured out is that the vendors who took my $300 order seriously are the same vendors I now call for $12,000 orders. And the ones who made it clear that two units was a nuisance — they don't get the $12,000 order, because I remember. Small doesn't mean unimportant. It means potential.

What It Actually Cost Us

This is the section I should've built two years earlier. Because the purchase price was never the number that mattered.

Downtime doesn't show up on the PO

Let me run one of ours. A conveyor motor failure in 2023:

• Line idle roughly 6 hours (~$180/hr in idle labor and lost throughput): ~$1,080
• Replacement motor: $215
• Expedited freight (2-day instead of 7): $180
• Maintenance overtime to install after shift: $220
• Rough total: ~$1,695

Now here's the part that stings. The motor I should have bought was about $60 more than the one I did buy. I saved $60 and spent $1,695. And that's a cheap failure — the ones that hit during a committed delivery week cost considerably more than that in credibility, which brings me to the next thing.

Expedited freight is a tax on bad planning, and it's invisible

I went back through our records. Roughly 40% of my motor orders in 2023 were rushed. Rush premiums on those ran somewhere between 30% and 60% over standard freight and lead time. Add it up and it's a low-four-figure line item that never got approved by anyone, because it never appeared as its own budget line. It just dissolved into 40 different invoices.

(These numbers are from our own records and our own shop — yours will differ. The pattern probably won't.)

The invoice finance rejected

In 2022, I found a marketplace seller listing two motors about $190 cheaper than our regular source. Fast shipping, good reviews. I ordered. They arrived quickly and they were the right frame size.

What they didn't arrive with was a proper invoice. Handwritten receipt, no tax ID, no PO reference number. Finance rejected the expense report outright and there was no path to appeal it. I ate $1,900 out of the department budget, and then I had to explain to my operations manager why the line had been down for a week over a $190 savings.

I verify invoicing capability before I place any order now. Before price. Before lead time. That's the first question.

The cost that never appears on a report

When I tell our operations manager "it'll be here Thursday" and it isn't, I'm not the person who bought a bad motor. I'm the person whose Thursday estimate can't be counted on. That one doesn't reconcile against anything. It just quietly changes how much room I get to make my own calls.

What Changed

The fix was not exotic, which annoyed me a little.

• Every motor request now starts with a short written form: frame size, mounting, voltage, phase, poles or step angle, load description, ambient conditions, duty cycle, and whether the motor is drive-fed.
• I ask the vendor to confirm the specification back to me in writing before I issue the PO. The ones who can't do that are telling me something useful.
• I stopped splitting the motor and the drive across two vendors. One throat to choke.
• I stopped assuming a large manufacturer won't talk to me about a two-unit order.

On that last point — I'd assumed for years that going direct or through an authorized distributor meant getting routed to a call center. It doesn't have to. The manufacturers who've been at this for decades tend to have published frame-size charts, published spec sheets, and distributor networks that will actually answer a question about a 2 hp motor destined for a 70-person shop.

TECO Electric & Machinery has been building motors since 1956. The reason that history matters to me isn't the logo on the nameplate — it's that a company with that track record has documentation I can hand to finance and to our maintenance lead, and a spec sheet that doesn't require me to translate marketing language. Searching for "teco electric and machinery" plus a frame size gets me to an actual datasheet. That sounds like a low bar. After two years of guessing, it's the bar that mattered most.

The Short Version

We didn't have a motor problem. We had a purchasing problem that manifested as motor failures, and it took me three expensive breakdowns and one rejected expense report to notice.

Once we started buying motors as components of a system — written spec, one accountable supplier, someone willing to answer questions on a small order — the failures mostly stopped. Not because we found better motors. Because we stopped guessing and started specifying.

If you're ordering motors the way I used to, from a part number in an email and the lowest quote on the page, you'll probably get away with it for a while. Then you won't, and the invoice won't explain why.