Shimano STEPS Power Transmission OEM logo
Technical article

Shimano STEPS E5000 Torque Nm: A Quality Inspector's Look at Torque, Bed-In, and Motor Types

2026-09-02 / Engineering Desk

Last March, our warehouse manager flagged a returned DU-E5000. The OEM's note said 'low assist — motor weak.' I'm the quality manager for a distributor that supplies Shimano STEPS drive units to e-bike builders. I review about 200 units a year, and in 2024 I rejected 6% of first deliveries for torque or brake-drag problems. That return looked like a torque problem. Here's the thing: it wasn't.

I put the bike on a roller dyno and measured rear-wheel output, then converted it back to an equivalent torque at the crank. The Shimano Steps E5000 torque Nm spec, according to the current Shimano STEPS product pages (shimano-steps.com, accessed March 2025), is 40 Nm at the drive unit. Our first reading equated to 25 Nm. Second reading: 26 Nm. The data said replace the motor. My gut said look at the brakes.

The numbers said replace the motor. My gut said look at the brakes.

Why did my gut push back? I spun the rear wheel by hand before strapping the bike down. It didn't spin freely. A healthy rear wheel on a Shimano STEPS bike should have a little drag from the gearbox, but this felt like a brake was on. The bike had brand-new disc rotors and pads. The pad surfaces had a shiny, glassy look. Bedded-in Shimano pads normally have a matte, even transfer layer. That was glazing.

Why the torque reading dropped

Brake drag eats torque. The motor was producing drive torque, but the rear brake was resisting the wheel. The dyno measures net output after that resistance, not the motor's gross output. So a bike with a dragging rear brake can read far below spec even when the motor is healthy. In this case, the brake was consuming roughly 15 Nm equivalent at the crank. The motor wasn't weak. It was being held back.

We removed the rear caliper, re-centered it, and bedded the pads in. After that, the same unit measured 39.8 Nm equivalent. The motor was fine.

What's a stepper motor?

One of the technicians asked what kind of motor is inside the STEPS unit. Is it a stepper motor? That question comes up more than you'd think, especially from people who work with 3D printers and CNC machines.

A stepper motor is a brushless DC motor that moves in fixed angular increments. The controller sends pulses, and the motor rotates a set number of steps per pulse. It's usually an open-loop design, which means the controller doesn't know if the motor actually moved. That works for positioning a print head; it would be a poor choice for an e-bike drive unit.

The STEPS motor is closer to a brushless DC servo motor. It has rotor position sensors, torque sensors, and speed sensors. The controller continuously adjusts current to match pedal input. If you're from the automation world, you can describe the DU-E5000 as a brushless DC servo motor with a crank torque sensor, and that's not wrong. It is not a stepper motor.

Shimano disc brake bed-in procedure steps

Here's the part that saved us. The Shimano disc brake bed-in procedure steps we use are simple, and they're not optional when you care about torque readings.

  1. Clean the rotor with isopropyl alcohol. Check the pads for oil, shipping wax, or glazing.
  2. Center the caliper. Squeeze and hold the brake lever, then tighten the caliper bolts.
  3. Ride on a flat, quiet stretch at about 10 to 12 mph.
  4. Pull the brake lever progressively. Slow down until you're almost stopping, but don't lock the wheel.
  5. Let the bike roll and repeat 15 to 20 times.
  6. Do 5 to 10 harder stops from 15 to 18 mph.
  7. Let the brakes cool completely before normal riding or a torque test.

Bed-in is not just for new pads. It's for new rotors too. If you install a rotor with old pads, the pad edges have to form a clean contact patch against the thicker rotor. Same basic steps, just slower and gentler.

One local service center skipped this procedure to save time. They installed cheap pads, didn't bed them in, and the customer came back a month later with a grooved rotor. The $15 pad saving turned into a $90 rotor bill. I do not mean the pad was the only problem. I mean the missing bed-in made it a problem.

What does 40 Nm actually feel like?

After the fix, I rode the E5000 on our test loop. On flat ground, it feels like a strong tailwind. On a 6% grade, it will keep you moving at a sensible pace if you shift early. It is not the motor for a loaded cargo bike or a steep mountain shuttle. The STEPS line has higher-torque options for that. The E5000 is meant for urban commuting and light hills, and that's exactly where it belongs.

The 40 Nm peak doesn't tell you about the assistance curve. The E5000 ramps power from the torque sensor smoothly. That curve matters more than the peak number. A motor with the same peak but an on/off switch feels dangerous in a parking lot. Quality control should test the curve, not just the peak.

What about 5th gear motors?

Every few months, someone asks if they can make an E5000 feel faster by adding a 5th gear motors hub. A 5th gear motors hub spins the wheel directly, and it can feel aggressive because it pulls without needing to shift. But a crank drive applies torque through the chainring, so the rider can use the bike's gears to multiply force. The E5000's 40 Nm is rated at the crank, not at the motor shaft. That's why a 40 Nm crank drive can climb a steeper hill than a high-speed hub with a higher raw motor number.

I'm not going to pretend I know every hub motor on the market. If a customer wants a flat-ground speed machine, there are competent 5th gear motors builders who can do that. But for an OEM building a torque-sensing pedal-assist bike, the STEPS system is a different animal. Don't compare them by peak power alone.

Three lessons from a 40 Nm 'failure'

We sent that unit back to the OEM with a note: motor passed, brake failed. They added the bed-in procedure to their dealer checklist, and returns on that model dropped by a third the next quarter.

Here's what I'd tell any specifier:

  1. Verify the brake before blaming the motor. Torque readings lie when the brake drags.
  2. Read the torque spec correctly. The Shimano Steps E5000 torque Nm number is a maximum at the crank, not a continuous average. It's not weak; it's designed for urban commuting and light hills.
  3. Know the motor type. It's not a stepper motor. It's a closed-loop brushless DC servo motor with gear reduction. If that distinction doesn't matter to you, fine. If it matters to your customers, choose the system that does it properly.

I still review 200 units a year, and I still expect the occasional surprise. But now I check brake bed-in before I pull a motor, and I sleep better because of it.

Shimano STEPS Engineering Desk

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