Needle roller bearing detail for article header
Engineering Note

Single-Phase AC, Brushless DC, or Stepper? A Purchasing Guide for People Who Actually Have to Place the Order

2026-11-17 by Ingrid Nystrom

When I took over purchasing for our maintenance and fabrication group back in 2021, I assumed buying a motor was a spec-matching exercise. Find the horsepower, find the frame size, place the order. Done.

Three years, one shelf of wrong-sized gearboxes, and a very patient maintenance lead later, I know better. The right answer depends entirely on what the machine is actually doing—and the people taking the order rarely ask.

So here's how I break it down now. Four scenarios. Find yours.

The four situations, short version

  • Scenario 1: Steady load, single-phase power available, nobody cares about speed control.
  • Scenario 2: Speed has to change mid-run, and stopping position matters.
  • Scenario 3: Repeatable indexing on a tight budget, and you can live without feedback.
  • Scenario 4: You're replacing something that already exists. This is the one most people get wrong.

Scenario 1: Steady load, single-phase power, no speed control

This is the boring case, and boring is good. A single phase ac motor paired with a worm gear reducer covers a surprising amount of what a plant actually needs—conveyors, mixers, door operators, small pumps.

Two things to check before you order anything. First, the duty cycle. If the motor's going to run continuously, you want continuous-duty rated; a lot of cheap units are rated for intermittent duty and the label doesn't say so in large enough letters. Second, the service factor. A 1.15 service factor is the common baseline for general purpose work, and you should not treat it as headroom you're entitled to spend on a jam-prone machine.

Cost reality, from our own POs in Q3–Q4 2024 plus publicly listed distributor pricing: a ½ to 1 HP single-phase motor runs roughly $180–$450 depending on frame and enclosure, and a matched worm reducer adds about $200–$600. Total of a few hundred to just over a thousand, installed.

One tradeoff worth naming out loud: worm gearing is quiet and gives you a compact right-angle package, but it's less efficient than helical. Common figures put worm reducers in the 50–90% range depending on ratio and size, versus roughly 95–99% for helical. On a 12-hour-a-day machine, that gap shows up on the electric bill eventually. On a two-minute-a-day door operator, it never will.

Scenario 2: Speed has to change, and stopping position matters

Now you're in brushless dc motors territory. Permanent magnet rotor, electronic commutation via a drive, no brushes to wear out. Compared with an equivalent brushed DC unit you get more torque per frame size and a much longer service life—the wear parts moved into the drive, where they mostly consist of capacitors.

The catch nobody mentions at quote time: the drive is now part of your equipment, and somebody has to know how to parameterize it. We bought a brushless setup in 2023 for a packaging infeed, and the commissioning took a full day because none of us understood what the drive's gain settings were doing. The motor was fine. We weren't.

Budget roughly $400–$1,200 for the motor and $300–$900 for a drive in small industrial sizes, then add commissioning labor you probably didn't plan for. If you need closed-loop position control, add an encoder and another few hundred dollars.

And this is where the vendor relationship matters more than the spec sheet. One rep told me flat out: "We don't do washdown-rated enclosures for this line—here's who does it better than us." That one sentence earned him every other order I've placed since. A supplier who names the edge of their own capability is telling you the rest of what they say is probably true.

Scenario 3: Repeatable indexing on a budget

People ask what's a stepper motor like it's a trick question. It isn't. A stepper motor moves in discrete increments—a common size takes 200 full steps per revolution, so 1.8 mechanical degrees per step—and it does that on command from a pulse train. No encoder, no feedback loop, no confirmation that the shaft actually went where you told it.

That open-loop nature is the whole value proposition and the whole risk. For a label applicator, a small indexing table, or a dosing pump that never fights a hard load, a stepper is genuinely the cheapest sane answer. For anything where the mechanism can jam or the load can suddenly spike, it'll quietly lose steps and your machine will drift a little further off home every cycle until somebody notices a hundred parts later.

Here's the part I got wrong. I bought a stepper in 2022 to save money over a servo package, on a station that occasionally binds up on misfeeds. It skipped steps maybe once every two hundred cycles. Nobody caught it for a week. The scrap cost exceeded what the servo would have charged us on day one. I don't have hard data on how often that pattern repeats across other plants, but anecdotally, the answer is more than people admit.

Steppers also hold position well when stationary—holding torque is a real spec—so they're not wrong for vertical or holding applications. Just size the motor against your real torque demand with at least 50% margin over what you calculated, not against the number you're hoping for.

Scenario 4: You're replacing something that already exists

This is where most of the expensive mistakes happen. Everybody starts by searching the exact model number off the nameplate, finds a similarly numbered unit, orders it, and then stands in the shop holding a gearbox that doesn't quite bolt up.

What you actually need to match, in this order: mounting interface (foot or flange pattern, bolt spacing, shaft height), shaft diameter and keyway, gear ratio, and output torque at your required speed. The part number is the least useful of the five, because model lines get revised and suffixes change meaning.

Boston Gear's published 700 series catalog is genuinely useful here even if you're not buying their product—the dimensional drawings are complete enough to cross-reference against whatever you're pulling out. That's the advantage of an old catalog: the drawings predate the era of hiding dimensions behind a configurator.

And yes, if you're in the Charlotte area, there are distributors who stock or can overnight motion control components. Ask them upfront whether the unit ships from their shelf or from a factory—those are two very different lead times and the website doesn't distinguish.

Now the counterintuitive part. Everyone wants a drop-in replacement. I'd push back on that instinct. Once a reducer has run for ten or fifteen years, the surrounding hardware has moved—mounting pads sag, couplings wear, shims compress. A perfectly dimensioned replacement can still need re-shimming, a new coupling, and a realignment. I skipped verifying keyway length once because the frame size matched exactly and what are the odds? The odds caught up with me at 6 a.m. on a Monday. Two days of downtime and a $400 overnight freight bill.

So budget for installation. It's usually four to eight hours of a millwright's time, and it's the difference between a fix and a second problem.

How to figure out which scenario you're in

Six questions, and you'll know:

  1. What power is actually at the panel? Single-phase only, or is three-phase available? This alone rules out half the catalog.
  2. Does speed need to change during operation, or only at setup? If it never changes once it's dialed in, don't pay for a drive.
  3. How tight is your stopping tolerance? Anything measured in thousandths needs feedback. Anything measured in "close enough" probably doesn't.
  4. How many starts per hour, and how long is the run? High cycling kills motors that would otherwise last a decade.
  5. What's the environment? Washdown, dust, ambient heat, vibration. Check the IP rating against IEC 60529 before you assume a standard enclosure is fine.
  6. What happens if it stops at the wrong moment? This is the real budget question. If the answer is "we lose a batch," buy feedback. If it's "someone walks over and restarts it," don't.

Honestly, I'm not sure why more suppliers don't just ask questions two and six before quoting. My best guess is that the order-taker and the application engineer are different people, and the order-taker gets measured on speed.

One small note on search results

If you landed here looking for baby gear in Boston—strollers, car seats, that sort of thing—this is not that page. Different gear entirely. I only mention it because I manage the web inquiries too, and the confusion comes up more often than you'd think.

For everyone else: check your power supply, check what happens when the machine stops unexpectedly, and don't insist on a drop-in. That's most of the battle.

Share this engineering noteDiscuss with a spec engineer
Ingrid Nystrom

Ingrid Nystrom

Ingrid Nystrom is an independent rotating equipment reliability analyst covering gear units, bearings, couplings, lubrication, condition monitoring, and failure investigation. She uses ISO 20816-9 vibration evaluation for gear units alongside temperature, oil-debris, noise, alignment, and trend-baseline measurements to distinguish developing faults from normal variation. Her maintenance articles help reliability teams set alert limits, prioritize inspections, plan corrective work, and compare lifecycle consequences before downtime becomes critical.