There is a stepper motor on a lab syringe pump that stalls at exactly 3.2 rev/s. It runs fine at 2 rev/s, stalls at 3.2, and runs fine again at 4.5. The driver is a standard chopper drive running at 48V, the motor is an NEMA 23 with 2 Nm holding torque. The customer swapped the motor, swapped the driver, even swapped the cable. Same stall at 3.2 rev/s every time. The motor is not defective. It is hitting mid-band resonance, and no amount of component swapping fixes that.
The first time this happens to a machine builder it looks like a mystery. The motor runs, then loses torque, then runs again. The stall is repeatable to the decimal. It happens with the motor loaded or unloaded. Sometimes the motor just buzzes in place without moving. The fix is not a bigger motor. A bigger motor moves the resonance to a different speed, which means the problem shows up at a different point of the application range. The fix is damping, not horsepower.
Where the resonance comes from
A stepper motor is a two-phase motor driven in steps. Between steps, the rotor settles into a magnetic equilibrium position. If you apply a small torque and release it, the rotor oscillates around that position like a pendulum. The natural frequency of that oscillation depends on the rotor inertia and the holding torque. For a typical NEMA 23 with a 500 g·cm² load, that natural frequency sits between 2 and 5 rev/s of step frequency.
When the step rate approaches the natural frequency, each step kicks the rotor at exactly the wrong time. The oscillation grows instead of decaying. The rotor overshoots the target position, comes back, overshoots the other way. At some point the rotor is 180 degrees out of phase with the drive pulses. It stalls. The motor is still energized, still drawing current, still making the classic hum. It just is not moving.
The math is worth doing once. Natural frequency in steps per second:
f = (1 / 2π) × √(T_h / J)
Where T_h is holding torque in N·m and J is total inertia in kg·m². For a 2 Nm motor with 0.001 kg·m² total inertia (rotor plus load), f ≈ 225 steps/s for a 200-step motor. In rev/s that is about 1.1 rev/s, which is the resonance at half-step. Full-step resonance lands higher. The 3.2 rev/s stall on the syringe pump was the second-order resonance — the natural frequency and its harmonics. The damping ratio of a stepper is low, around 0.05, so the resonance peak is tall and sharp.
The fixes that actually work
Component swapping does nothing because the resonance is a property of the motor-inertia system, not a fault in one part. The fixes that work, in order of preference:
1. Microstepping. Running the motor at 8 or 16 microsteps per full step spreads the torque pulses out. The excitation at the resonance frequency drops. A motor that stalls at full-step will often run clean at 8 microsteps. The syringe pump was switched to 16 microsteps and the stall disappeared. Cost: zero, it is a DIP switch on the driver.
2. Increase the damping. A damper (a small inertia ring with a rubber element) bolted to the shaft absorbs the oscillation energy. It adds inertia, which moves the resonance down, and adds damping, which flattens the peak. A $25 damper fixed a machine that had eaten two “good” motors. The downside is slower acceleration, acceptable for most index moves.
3. Adjust the current reduction. Some drivers let you set a current reduction when the motor is idle. If the reduction happens while the motor is stepping through the resonance band, the torque drops and the stall gets worse. Set idle current reduction to off, or set it high enough that the resonance band is crossed before the reduction engages.
4. Skip the resonance band. If the application runs at a fixed speed, program the motion profile to accelerate through the band quickly. The motor crosses 3.2 rev/s in 100 ms instead of dwelling there. The syringe pump’s application was fixed-speed, so this alone would have worked. The customer combined it with microstepping anyway.
One more option: a closed-loop stepper with an encoder. The encoder watches the rotor position. If the rotor starts to lose sync, the drive applies corrective current in real time. Closed-loop steppers eliminate the resonance stall almost completely. They cost about 30% more than open-loop. For machines that must not stall, they are the right answer.
Where the resonance shows up in practice
The classic complaint is “my motor stalls at a certain speed and I can’t find anything wrong.” That complaint is the diagnosis. The motor is fine, the driver is fine, the wiring is fine. The system has a natural frequency and the application happens to run at it. The list of things to check, in order:
- Does the stall happen at a fixed step frequency regardless of load? Resonance. If it moves with load, it is a torque problem.
- Does the motor buzz and vibrate hard at the stall speed? Resonance. A torque stall runs hot and stops cleanly.
- Does microstepping fix it? Resonance. It almost always does.
- Does a damper fix it? Resonance. The damper is the mechanical cure.
A torque problem gets worse with load and heat. A resonance problem is repeatable, speed-dependent, and load-insensitive. That one distinction separates a component problem from a system problem, and it saves a lot of pointless swapping.
Designing it out before it happens
On new designs, the resonance is predictable. Calculate the natural frequency at the design stage. If the operating speed band overlaps the resonance, change one of three things before building: pick a motor with different inertia (add a flywheel or damper), use microstepping from the start, or design the motion profile to accelerate through the band. The last one costs nothing and is usually enough.
For existing machines with an unexplained mid-band stall, the fix order is microstepping first, damper second, closed-loop third. All three are cheaper than the troubleshooting time already spent swapping parts. The lab pump customer’s total cost to fix: one hour of programming (microstep setting and a profile change) and zero parts. The two weeks of swapping had cost more than that in labor alone.
A stepper that stalls at a fixed speed, buzzes hard, and is load-insensitive is hitting resonance, not failing. Microstep it, damp it, or skip the band in the motion profile. Do not buy another motor. The motor you have was never the problem.
