Explainer
Why a Stepper Controller Cannot Drive a Servo Motor
Short answer
No, not directly. A stepper controller sequences phase currents open loop, while a servo drive commutates against encoder or resolver feedback; the only bridge is a servo drive that documents a step/direction input. A servo motor that hums without moving has only proven it was energized, not that the command or the loop is compatible.
On this page
- What is the difference between a stepper controller and a servo controller?
- What signals does a stepper system actually use?
- What command inputs does a servo drive accept?
- Why does the servo motor hum but not move?
- Can a stepper controller run a servo motor?
- How fast do step pulses need to be for a servo axis?
- What do I check before wiring step and direction into a servo drive?
- What is the cheapest way to get an old servo machine running?
- Questions asked
No, not motor to drive. A stepper controller, really a stepper drive with a translator, switches phase currents in an open-loop sequence and expects the motor to follow; a servo motor is a commutated machine whose drive measures rotor position with an encoder or resolver and closes a loop on it. The only bridge is the command format: many servo drives also accept step and direction pulses, and when they do, your pulse source commands the servo drive, and the drive still closes the loop. A servo motor that only hums on a stepper drive has been energized, nothing more; that is not a compatibility test. This article separates the controller, drive, and motor, compares the interfaces, and gives the wiring checks and the decision path for making an old machine run.
What is the difference between a stepper controller and a servo controller? #
A stepper controller sends open-loop position commands, step pulses and a direction level, to a drive that energizes motor phases in sequence; a servo system commands a drive that measures rotor position with an encoder or resolver and closes a torque, speed, or position loop on that feedback.
The difference that decides compatibility is where position information lives. In a stepper system, position lives in the pulse count leaving the controller: the motor follows blind, and an overload makes it lose position silently while the controller keeps counting. In a servo system, position lives on the motor shaft, and the drive continuously drives the error between commanded and measured position toward zero.
One word causes most of the confusion: controller. There are two boxes answering to that name. The motion controller, a CNC program on a PC, a PLC motion module, or a dedicated card, generates the command. The drive is the power amplifier connected to the motor. Small stepper packages merge the two into one box, which is why a stepper controller gets wired straight to motor leads. Servo systems keep them separate, and the drive additionally needs the motor's feedback wired back into it.
What signals does a stepper system actually use? #
Two digital lines: a step pulse train and a direction level. Each step pulse advances the motor one step or microstep, 200 full steps per revolution on the common 1.8 degree hybrid two-phase stepper, and the drive's translator converts that into the phase-current sequence.
The step line carries one pulse per increment and the direction line holds a logic level for rotation sense; this two-signal scheme driving a two-phase bipolar stepper is the case every stepper application note describes. Microstepping scales the phase currents to split each full step into fractions.
Nothing comes back. The controller never learns rotor position, phase current, or torque, and the drive's power stage is built for the stepper's two phase windings at their rated current. It has no commutation reference because none is needed.
What command inputs does a servo drive accept? #
Three command families exist: a plus/minus 10 V analog signal commanding velocity or torque, a digital fieldbus or PWM interface, and, on many servo drives, the same step and direction pulses a stepper uses, with the drive closing its own loop on motor feedback.
The command input only sets the target; the feedback is what makes it a servo. Every servo drive expects its own motor's encoder or resolver on a dedicated connector, and most will not leave the disabled state until that feedback is present and configured.
| Command interface | What it carries | Pairing with a pulse-only controller |
|---|---|---|
| Step and direction | Position increments on two digital lines | Direct, but only when the servo drive documents this input |
| Plus/minus 10 V analog | Speed or torque, sign or a digital input gives direction | Needs a converter or a controller with an analog output |
| Digital network | Position, speed, and torque over one fieldbus | Needs a controller that speaks the same protocol |
The spread is real and documented. The B8501 brushless analog position-control manual states that its step and direction input can command position, velocity, or torque, while its analog input commands velocity or torque. Direction can be handled more literally still: on the A6-RS series, one of the eight DI input ports can be set to function 17 (Speed Command Direction) to steer an analog speed command. The drive manual, not the motor type, decides what you wire.
Loop tuning also lives in the drive, and it is a real adjustment: gains set too high for the load make the axis hunt, the failure mode covered in why a servo motor oscillates between two positions.
Why does the servo motor hum but not move? #
The hum proves only that the power stage energized the windings; it proves nothing about the command format, the commutation, or the control loop. Current without properly sequenced commutation produces buzz and heat, not average torque in one direction.
Two different wiring mistakes produce exactly that symptom, and they have different fixes.
If the stepper drive's phase outputs were connected to the servo motor's power leads, the drive is holding chopped DC in two windings with no reference to rotor position. A permanent-magnet synchronous motor fed unsynchronized current makes torque ripple that averages near zero: the rotor buzzes against its magnetic alignment and the windings heat. That is a drive-topology mismatch, and no command wiring fixes it.
If the pulses reached a servo drive's command input and the motor still only hums, work the cheap checks first: read the drive's fault indicator or software for a feedback or enable fault; confirm signal common between the pulse source and the drive; verify pulse levels and frequency against the drive's documented input; then confirm the motor's encoder is plugged into the same drive. Only after the drive enables cleanly and still refuses to move is the motor itself the suspect, and how to tell if a servo motor is actually faulty settles it on the bench.
Can a stepper controller run a servo motor? #
Yes, through the command side only: feed the stepper controller's step and direction pulses into a servo drive that documents a step/direction input, connect that drive's motor and encoder, and the servo runs as a closed-loop follower of your pulse count.
What changes is where the loop lives. The drive counts incoming pulses as the position target and commutates the motor from its own feedback to reach that count. The controller still only pulses; commutation, correction, and stall detection stay in the drive. A stall trips a following-error fault instead of silently losing position, which is the practical gain over a stepper fed the same pulses.
The trade is headroom: the pulse rate caps speed at your chosen resolution, which is the arithmetic in the next section. The bridge is command-side only; there is no power-side equivalent, so a stepper drive never commutates a servo motor regardless of what pulses you feed it.
How fast do step pulses need to be for a servo axis? #
Pulse rate equals resolution times shaft speed: pulses per revolution multiplied by revolutions per second. In a calculated example, a drive geared to 2000 pulses/rev at 3000 rpm needs 2000 × 50 = 100,000 pulses/s = 100 kHz.
A servo drive in step/direction mode has a configurable command resolution, the counts per revolution it expects, set in the drive rather than taken from the motor. That setting is the multiplier for every speed the axis will ever reach.
Calculated example for these inputs:
- Command resolution: 2000 pulses/rev (drive setting)
- Target speed: 3000 rpm = 3000 ÷ 60 = 50 rev/s
- Required pulse rate: 2000 pulses/rev × 50 rev/s = 100,000 pulses/s = 100 kHz
Pulse sources have ceilings you must check rather than assume, and sizing a PC to output 25 kHz step pulses is its own constraint on older PC-based controllers, so put your controller's maximum frequency into the arithmetic before committing to a resolution. If that source tops out at 25 kHz: 25,000 ÷ 2000 = 12.5 rev/s = 750 rpm, far short of the target. Re-geared to 400 pulses/rev, the same source reaches 25,000 ÷ 400 = 62.5 rev/s = 3750 rpm, above the 3000 rpm target, at the cost of coarser resolution, 360° ÷ 400 = 0.9° of shaft per pulse versus 0.18° at 2000.
Decision for this example: gear the drive to 400 pulses/rev rather than chase resolution the mechanics cannot hold, or move to an analog or networked command if the resolution genuinely matters.
What do I check before wiring step and direction into a servo drive? #
Five checks, in rising order of cost: the drive's documented command mode, the pulse signal levels and minimum pulse width, the enable and fault state, the motor's feedback wired to its own drive, and a small following-error limit set before the first motion.
- Command mode. Set the drive's command-source selection to step and direction; on drives without a dedicated switch, this lives among the command parameters, and the manual names it.
- Signal levels. 5 V logic versus 24 V, optocoupler inputs and their current-limiting resistors, sinking versus sourcing, and the documented minimum pulse width and maximum frequency. Mismatched levels are the quietest failure: the drive sees nothing, or noise.
- Enable and faults. Assert the enable input and clear any fault; a drive held disabled by missing feedback will not respond to any pulse stream.
- Feedback pairing. The motor's encoder or resolver goes to its own drive; feedback borrowed from a neighboring axis makes the loop correct the wrong motor.
- Limits first. Set the following-error limit, and any velocity limit the drive offers, before the first commanded move.
One safety note before the rest: isolate the machine and give the servo drive's DC bus time to discharge before touching power or motor leads. Then make the first moves with the shaft uncoupled or the load blocked and an e-stop within reach, because a servo with a reversed direction or a mispaired feedback can spin to full speed immediately.
What is the cheapest way to get an old servo machine running? #
Keep the motors and replace the drives if the mechanics are sound: servo drives with a documented step/direction input are the least-rework path for a pulse-only controller, a converter or new controller covers analog drives, and only dead hardware justifies a stepper retrofit.
- Motors and drives healthy, controller pulses only: configure the existing drives' step/direction mode if the manual documents one. If it does not, replace the drives with units that do. The mechanics, wiring tray, and controller stay untouched.
- Drives analog-only: a step/direction-to-analog converter between the controller and each drive closes the gap, and interface converters built for exactly this retrofit work against the industry-standard plus/minus 10 V command, some alongside a tachometer input for older velocity-mode drives. The alternative is a motion controller with real analog outputs.
- Motors or drives dead: weigh a stepper retrofit on the same mechanics. The axis then holds position open loop and must be sized for peak torque and duty, the trade-offs laid out when choosing servo or stepper for a speed-regulated axis.
The command format is the only bridge between the two worlds; the feedback loop is the difference.
Questions asked
Will a step/direction servo drive still fault if the axis stalls?
Yes. Because the drive closes its loop on the motor's encoder, a stall shows up as following error: the pulse count commands motion, the encoder reports none, and the drive faults instead of losing position silently. That is the practical gain over a stepper, which stalls while the controller keeps counting. Set the following-error limit to the lag the axis may show during normal acceleration before you enable it.
What does a ±10 V analog command actually tell a servo drive?
Usually speed, sometimes torque. Zero volts means stop, the polarity sets direction, and 10 V is full command in that direction. Position is not in the signal at all: the drive regulates what it is told, and the position loop has to live upstream in the controller. That is why older analog systems take longer to configure and why a step/direction or networked drive is usually the easier retrofit.
My servo drive will not enable. Is the motor bad?
Probably not yet. Most servo drives refuse to enable when the motor's feedback is missing, misconfigured, or wired to a different drive, and the drive's fault indicator or software names that condition before anything else. Clear the feedback fault, then check the enable input's logic level. Only after the drive enables cleanly and still refuses to move under command should you suspect the motor itself.
How many pulses per revolution should I configure?
Match the resolution to what the mechanics can hold, not the maximum. A 5 mm lead screw at 2000 pulses per revolution resolves 0.0025 mm per pulse, finer than most frames and gear backlashes will ever repeat. Coarsening the command resolution raises the speed a given pulse rate can reach, so spend pulses only where the machine can actually use them.
Can I just swap the servo motors for stepper motors on the same mechanics?
If the axis can be sized for it, yes, and it is a legitimate decision rather than a downgrade by definition. Steppers hold position open loop and lose it silently on overload, so the peak torque demand, speed range, and duty cycle decide the answer. Compare the rework of drives-and-converter against a stepper retrofit before choosing, because the lighter rework usually wins on an older machine.