Shimano STEPS Power Transmission OEM logo
Technical article

Don't Order a 'Motor' Without a Spec Sheet: Shimano Steps, NEMA 17, and VFD Lessons

2026-08-31 / Engineering Desk

'Can I get a quote for a motor?'

That's how the order started. On paper, it looked like a no-brainer. The quote was 38% lower than the component we'd used before. The supplier said it was compatible. We approved the purchase order before lunch.

It wasn't compatible.

I'm a procurement lead at a 110-person electric mobility company. I manage a component budget of about $1.2M a year, I've negotiated with 40-plus vendors, and I've tracked every PO since 2019. The mistake cost us $1,800 in rework and three production days. The worst part? We skipped a five-minute spec check.

The problem we thought we had was price

When I first started in this role, I assumed a motor was a motor. If it had similar watts and similar dimensions, it would work. Three budget overruns later, I learned about total cost of ownership.

The word 'motor' covers wildly different parts. In e-bikes, a Shimano Steps drive unit is a system: motor, torque sensor, speed sensor, controller, and firmware. In automation, a NEMA 17 stepper motor is one piece of a motion train. In a factory, VFD-compatible motors have their own electrical and thermal ratings. They share the noun 'motor.' They don't share a spec sheet.

Take the Shimano Steps E5000. According to the Shimano STEPS product page (shimano-steps.com, accessed March 2025), the E5000's maximum torque is 40 Nm. That number isn't just marketing. It determines hill-climbing ability, battery current draw, and how the assist firmware behaves. If you are the person checking the Shimano Steps E5000 torque Nm before ordering, you're already doing the step that my team skipped. If you don't check, you could end up with a unit that delivers half the expected pull and then overheats on a test ride.

The same logic applies to the Shimano Steps speed sensor. It's a small part, easy to ignore. But the e-bike uses that sensor to calculate wheel speed and assist cutoff. On one build, we used a supplier-confirmed 'equivalent' speed sensor. The connector looked the same. But it sent a different pulse count per wheel revolution. The system cut assist at 18 km/h instead of the expected 25 km/h. The bike came back, we swapped in the correct Shimano Steps speed sensor, and the $15 sensor cost us $160 in freight and labor before we were done.

The deeper problem: 'motor' isn't a spec

I'm not an electrical engineer, so I can't walk you through winding theory. What I can tell you from a procurement perspective is that every one of these failures showed up in my cost tracking system as rework, expedite fees, or downtime.

Here's what happened on the automation side. A colleague asked for a 'nema17 stepper motor.' NEMA 17 is a frame standard: the faceplate is 1.7 inches by 1.7 inches. It doesn't tell you holding torque, rated current, shaft length, step angle, or wiring configuration. We ordered the right frame size and the wrong current rating. The motor got hot and the driver stalled.

When the supplier finally sent a stepper motor diagram, it showed a 4-wire bipolar setup. Our controller was wired for a 6-wire unipolar motor. We had to buy the correct motor and rewire the harness. The 'bargain' motor ended up being the most expensive one we tested that quarter.

The same confusion shows up when someone asks, 'what motors are compatible with VFD?' The responsible answer: for continuous applications, use inverter-duty motors designed for VFD operation. Per NEMA MG 1 Part 31, those motors are built to handle the voltage spikes and thermal stress that come from variable-frequency drives. A standard motor might run for a while, then fail when the insulation breaks down. The motor wasn't defective. The application was wrong.

I assumed 'same specifications' meant identical behavior across suppliers. Didn't verify. Turned out each supplier interpreted the ambiguous term differently. Now when I compare quotes, I require a datasheet before I look at a price.

What ignoring specs actually costs

Over the past six years, I've analyzed about $180,000 in cumulative component spending across e-bike and automation projects. Around 9% of that went to fixing spec mistakes.

Let me put real numbers on it. A motor with the wrong connector needs an engineer to diagnose it. That's three hours of debugging at $90 per hour. Then there's return shipping, restocking, and an expedited replacement. Add one day of delay on an assembly line, and a $200 part can turn into a $1,400 line item.

Another hard lesson: I knew I should pull the wiring diagram before ordering, but I thought 'what are the odds? It's basically the same as last time.' The odds caught up with me. The old sensor had a 30 mm cable. The replacement had a 150 mm cable. The connector fit. The routing didn't. That's a $400 mistake over a cable length that was in the datasheet all along.

In Q2 2024, I compared two suppliers for a replacement sensor. One quoted $12.00; the other quoted $8.50. I almost chose the cheaper one until I read the datasheets. The cheaper sensor needed a different mounting bracket, which added $35 and two weeks of lead time. The 'higher-priced' supplier was the lower total cost. That gap almost never shows up on a purchase order comparison.

There's also a compliance risk. If you're building e-bikes for the EU market, EN 15194 sets limits for power and speed. A Shimano Steps drive unit is designed as one certified system. Swapping a speed sensor for an unverified part can create a question mark around that certification. I'm not a compliance specialist, so check with your certification body before making that trade. From a budget view, the risk is never worth four dollars.

The math is simple: five minutes of verification beats five days of correction. Most of the failures I've tracked were visible in the spec sheet before any money changed hands. We just didn't look.

What I do now: a pre-order checklist

I'm not asking you to become a technical expert. I'm asking you to make the spec sheet part of your purchasing process.

  1. Ask for the full datasheet before the quote. If a supplier hesitates, treat it as a red flag.
  2. Match the operating parameters that matter. For Shimano Steps: drive unit model, torque Nm, speed sensor part number, firmware compatibility, and mounting kit.
  3. For a NEMA 17 stepper motor, check holding torque, rated current, shaft length, step angle, and wiring. Ask for a stepper motor diagram if the datasheet doesn't include one.
  4. For VFD applications, confirm the motor is marked inverter-duty or rated per NEMA MG-1 Part 31. Ask how the VFD will be configured, including voltage, frequency, and cable length.
  5. Don't place the PO until the specs are written down. A verbal 'it will work' is not a specification.

Since I built this checklist, spec-related rework dropped by about 40% in the following year. It's not exciting. It's just cheaper.

Bottom line: the cheapest motor is the one you can install once and forget. The expensive one is the one that needs engineering time, expedited shipping, and a return label. Whether you're buying a Shimano Steps drive unit, a nema17 stepper motor, or a VFD-rated motor, the spec sheet is the contract. Read it before the PO, not after the failure.

Shimano STEPS Engineering Desk

Application notes from drive unit, brake and service documentation teams.