Motion-Control Systems: Precision, Power and Performance: A Machine Tool Slide Drifted 0.003 Inches

A Machine Tool Slide Drifted 0.003 Inches — and Scrapped an Entire Production Run

The first sign of trouble at the source manufacturing plant came at 2:47 AM on a Tuesday. A CNC milling machine — one that had been reliably cutting aerospace bracket slots for three years — started producing parts that were 0.003 inches out of tolerance. Not much, by most standards. But for the titanium brackets destined for a commercial aircraft wing assembly, it was catastrophic.

By the time the morning shift discovered the problem, 340 brackets sat in the reject bin. The root cause? A degraded feedback transducer in the machine's motion control system had been slowly drifting for weeks, and nobody had caught it. The closed-loop positioning system thought it was hitting the mark. It wasn't.

Total cost: over 280,000 units of currency in scrapped material, rework, and delivery penalties.

the practitioner, the plant's motion control engineer, would later call it "the most expensive lesson in sensor calibration I've ever witnessed." But the real lesson ran deeper than a single transducer. It was a lesson about the entire architecture of motion control — from the controller generating commands, to the motor converting electrical energy into mechanical power, to the feedback sensor closing the loop.

This is the guide the practitioner wished he'd had from day one.

Whether you're designing a new motion control system, troubleshooting an existing one, or simply trying to understand why your machine does what it does (or doesn't), everything you need is here. Every control architecture. Every motor type. Every transducer, valve, and damping method. No shortcuts. No fluff.

Why Motion Control Is the Foundation of All Precision Manufacturing

The most important factor in the manufacture of accurately machined components is the control of motion, whatever power source is used. Every CNC machine, every robotic arm, every automated assembly station — they all depend on motion control systems that translate electrical commands into precise mechanical action.

For all practical purposes, motion control is accomplished by electrical or electronic circuits energizing or de-energizing actuators such as:

• Electric motors (rotary output)
• Solenoid valves connected to hydraulic or pneumatic cylinders (linear or rotary output)

The accuracy with which a machine tool slide may be brought to a required position, time after time, controls the dimensions of the part being machined. This accuracy is governed entirely by the design of the motion control system in use.

These systems span an enormous range — from milliwatts to megawatts of power output — and they control far more than just position. A motion control system may govern:

• Position (linear or rotary)
• Velocity
• Acceleration
• Combinations of these motion parameters
• Process variables such as voltage, tension, and force
• Solenoid-operated valves for fluid power systems

The three main factors governing the design of any control system are:

1. Whether the system is open-loop or closed-loop
2. What kinds and amounts of power are available
3. The function requirements of the application

Get any one of these wrong, and you're headed for trouble — the kind the practitioner discovered at 2:47 AM.

Control System Application Factors: The Design Checklist You Cannot Skip

Before you select a single component, you need to define the operating envelope. Here is the comprehensive checklist that experienced motion control engineers work through at the design stage:

💡 Visual Strategy: A decision-tree flowchart showing how each factor narrows the choice of control architecture (open vs. closed loop, motor type, feedback type) would be ideal here.

Skip any one of these factors, and you're designing blind. the practitioner's team had missed the ambient temperature factor — a slow seasonal temperature drift was shifting the transducer's calibration without triggering any alarm thresholds.

Open-Loop Systems: Simple, Affordable, and Dangerously Limited

The term open-loop typically describes the use of a rheostat or variable resistance to vary the input voltage and thereby adjust the speed of an electric motor. This is a low-accuracy control method because there is no output sensor to measure actual performance.

You set a command. The system executes it. Nobody checks whether it actually got there.

When Open-Loop Works

Open-loop is not always inadequate. Stepper motors in open-loop systems can make them very accurate. Shafts of stepper motors are turned through a fixed angle for every electrical pulse transmitted to them. The maximum pulse rate can be high, and the shaft can be coupled with step-down gear drives to form inexpensive, precise drive units with wide speed ranges.

When Open-Loop Fails

Open-loop systems are only as accurate as the input-versus-output requirement can be calibrated, including the effects of:

• Changes in line voltage
• Temperature variations
• Other operating conditions

Although average speed with stepper motors is exact, speed modulation can occur at low pulse rates, and drives can incur serious resonance problems. If your application cannot tolerate momentary velocity ripple or positional uncertainty under varying loads, open-loop alone won't cut it.

Closed-Loop Systems: The Gold Standard for Precision

In a closed-loop system, a command is input — by a human operator, another piece of system equipment, or a computer. The system response is converted by the output sensor to a compatible, scaled electrical signal that can be compared with the input command.

The difference constitutes an error signal.

That error is amplified and applied to a motor to produce a correction. When the amplifier gain is high (the difference is greatly enlarged), even a small error will generate a correction.

The Closed-Loop Architecture

                    ┌─────────────────────────────────┐
                    │                                 │
  Input     (+)     │   Error      ┌──────────┐      │    Output
  Command ──►(Σ)───►│──Signal────►│ Amplifier ├──►Driver──►Response
              ▲     │             └──────────┘      │
              │     │                                 │
              │     └─────────────────────────────────┘
              │
              │         ┌──────────────────┐
              └─────────┤ Feedback Sensor  │◄─────────
                        └──────────────────┘

Accuracy of closed-loop systems is directly related to the accuracy of the sensor. Choosing between open-loop and closed-loop controls often means choosing between low price and consistent, accurate repeatability.

Closed-Loop System Parameters

Every closed-loop system is defined by a set of critical performance characteristics:

The Gain Trap

Here's the paradox every controls engineer faces: high gain produces high accuracy, but high gain can also produce instability.

A high gain can lead to an unstable system due to inherent delays between electrical inputs and outputs, especially with the motor. Response accuracy depends not only on the precision of the feedback sensor and the gain of the amplifier, but also on the rate at which the command signal changes. The ability of the control system to follow rapidly changing inputs is naturally limited by the maximum motor speed and acceleration.

This is precisely why damping methods exist — and why understanding them separates competent engineers from those who create oscillating nightmares.

Damping Methods: How to Tame an Unstable System

When the practitioner's team finally stabilized their replacement system, the fix wasn't just a new transducer. They had to re-tune the entire damping architecture. Here are the methods available to you:

The best damping methods permit high error amplification and accuracy, combined with the desired degree of stability. Whatever form the output takes, it must be converted by the output sensor to an electrical signal of compatible form for comparison with the input command.

Control Motor Types: Matching the Motor to the Mission

Not all motors are created equal — and choosing wrong here is one of the most expensive mistakes you can make. Here is the complete classification of control motors and what each brings to the table:

AC Motors

DC Motors

The Peak Torque Advantage of Rare-Earth Magnets

An important feature of high-performance permanent-magnet motors using high-energy rare-earth magnets (samarium cobalt or neodymium) is that their maximum torque output can be 10 to 20 or more times higher than their rated torque.

In intermittent or low-duty-cycle applications, very high torque loads can be driven by a given motor. However, when rare-earth magnets are not used, peak torque capability may be limited by the possibility of demagnetization. Always verify peak torque capabilities for lower-cost motors using weaker Alnico or ferrite magnets.

Key Motor Parameters

For permanent-magnet motors, the most important characteristics are defined by relatively few parameters:

These parameters allow you to predict motor behavior under virtually any operating condition — from stall to no-load, from continuous duty to intermittent peak demands.

Electric Motors for Control Systems: Open-Loop vs. Closed-Loop Requirements

Motors used in open-loop systems (excluding stepper motors) need not respond quickly to input command changes. Where the command is set by a human, response times of hundreds of milliseconds to several seconds may be acceptable. Slow response does not lead to the instabilities that time delays can introduce into closed-loop systems.

Closed-loop systems need motors with fast response. The best are permanent-magnet DC units, used where wide bandwidth, efficient operation, and high power output are required.

💡 Visual Strategy: An idealized control motor characteristics diagram showing the relationship between speed (rad/sec) and torque (in.-lb), with key operating points labeled ($T_{STo}$, $T_L$, $\omega_{NL}$, $\omega_R$), would be highly valuable here.

Controllers: The Brain of the Motion Control System

The controller is the fundamental component that determines system versatility. Its primary characteristic is power output, which must be compatible with motor and load requirements.

Controller Design Requirements

When driving a DC motor, the controller must:

• Supply sufficient power to match load requirements plus motor operating losses
• Operate at minimum line voltage and maximum ambient temperature
• Ensure wiring is not greatly sensitive to transient or steady-state electrical interference
• Separate power lines from control signal lines (or shield and isolate to avoid cross-coupling)
• Include electrical interference filters on main lines so the control system does not affect the power source

Special Features of Controllers

Current Limiting vs. Voltage Limiting

Current limiting is more useful than voltage limiting. Electronic torque limiting via current control can frequently avoid the need for mechanical torque limiting (such as slip clutches to avoid overtravel damage). In many geared systems, most of the kinetic energy is in the motor — uncontrolled stops can be destructive.

Voltage limiting may be needed to:

• Isolate the motor from voltage transients on the power line
• Prevent overspeeding
• Protect electronic components

Smoothing Input Commands

An abruptly applied step command can be smoothed out so that heavy motor inrush currents are avoided. The penalty is a corresponding delay in response — but in many industrial applications, the trade-off is worthwhile.

Control Function: Defining What the System Must Actually Do

The function of the control is set by the equipment designer and needs careful definition because it is the basis for the overall design.

For instance, in positioning a machine tool table, you must consider:

• Speed of movement and permissible variations in speed
• Accuracy of positioning
• Repeatability
• Overshoot tolerances
• Dozens of additional factors specific to the application

This is where the practitioner's team first went wrong. They had specified the positioning accuracy but not the long-term drift tolerance of the feedback sensor under varying thermal conditions. The control function definition was incomplete.

Electromechanical Control Systems: Wiring Is Versatility

Wiring is the simplest way to connect components, so electromechanical controls are more versatile than pure hydraulic or pneumatic controls. The key to this versatility is often in the controller — changes to computer chips or software can usually change system performance to suit the application.

Wiring and Interference Management

Complex electromechanical systems require more knowledge of design and debugging than strictly mechanical systems. Critical wiring practices include:

• Separating power lines from control signal lines
• Appropriate shielding and isolation to avoid cross-coupling
• Interference filters on main lines to prevent the control system from affecting the power source (which may influence other equipment connected to the same source)

Mechanical Stiffness: The Foundation of Dynamic Performance

When output motion must respond to a rapidly changing input command, the control system must have a wide bandwidth. Where the load mass (in linear motion) or the polar moment of inertia (in rotary systems) is high, there is a possibility of resonant oscillations.

The Rules of Stiffness

For the most stable and reliable systems, with a defined load, a high system mechanical stiffness is preferred. To attain this stiffness requires:

• Strengthening shafts
• Preloading bearings
• Minimizing free play or backlash

In the best-performing systems, motor and load are coupled without intervening compliant members. Even tightly bolted couplings can introduce compliant oscillations resulting from extremely minute slippages caused by load motions.

Backlash: The Hidden Compliance

Backlash is a factor in the effective compliance of any coupling but has little effect on the resonant frequency because little energy is exchanged as the load is moved through the backlash region. However, even in the absence of significant torsional resonance, a high-gain control system can "buzz" in the backlash region. Friction is often sufficient to eliminate this small-amplitude, high-frequency component.

The Direct-Drive Dilemma

The difficulty with direct-drive control systems lies in matching motor to load. Most electric motors deliver rated power at higher speeds than the driven load requires, so load power must be delivered by the motor operating at a slow and relatively inefficient speed.

Shaft power at low speed involves correspondingly high torque, which requires:

• A large motor
• A high-power controller
• Acceptance of high motor copper losses (heating)

However, direct-drive motors provide maximum load velocity and acceleration, and can position massive loads within:

• Seconds of arc (rotational)
• Tenths of thousandths of an inch (linear)

...under dynamic conditions.

The Geared Alternative

Where performance requirements are moderate, the required load torque can be traded off against speed using a speed-changing transmission — typically a gear train. The transmission effectively matches the best operating region of the motor to the required operating region of the load, and both motor and controller can be much smaller than would be needed for a comparable direct drive.

Torsional Vibration: The Invisible System Killer

Control system instabilities can result from insufficient stiffness between the motor and the inertia of the driven load. The behavior is similar to a torsional pendulum, easily excited by commanded motions of the control system. If frictional losses are moderate to low, sustained oscillations will occur.

The Resonant Frequency Formula

Despite the complex dynamics of the closed-loop system, the resonant frequency is given to a high degree of accuracy by:

$$f_n = \frac{1}{2\pi} \times \sqrt{\frac{K}{J_L}}$$

Where:

• $f_n$ = natural frequency in hertz (Hz)
• $K$ = torsional stiffness in in.-lb/rad
• $J_L$ = load inertia in in.-lb-sec²/rad

If this resonant frequency falls within the bandwidth of the control system, self-sustained oscillations are likely to occur. These oscillations are often overlooked by control systems analysts because they do not appear in simple control systems — and they are very difficult to correct.

Why Friction Is a Double-Edged Sword

Friction inherently reduces oscillation by dissipating energy in the system inertia. Coulomb friction (opposing motion but independent of speed) is especially effective in damping out oscillations when there is backlash between motor and load.

However, the required friction for satisfactory damping can be excessive, introducing:

• Positioning error
• Additional motor and controller power requirements
• Variability with operating conditions and time

The Notch Filter Solution

The most common method of eliminating torsional oscillation is to introduce a filter in the error channel to shape the gain characteristic as a function of frequency.

The simplest shaping network is the notch network — a band-rejection filter that sharply reduces gain at the notch frequency. By locating the notch frequency to balance out the torsional resonance peak, the oscillation can be eliminated.

Critical limitation: If the torsional resonance is within the required system bandwidth, little can be done except stiffening the mechanical system and increasing the resonant frequency. If the filter reduces the gain within the required bandwidth, it will reduce performance. This method only works if the natural resonance is above the minimum required performance bandwidth.

Where there are several modes of oscillation, several filter networks can be connected in series.

Stepper Motors: Precision Without Feedback (When Properly Applied)

In a stepper motor, power is applied to a wound stator, causing the brushless rotor to change position to correspond with the internal magnetic field. The rotor maintains its position relative to the internal magnetic field at all times.

How Stepper Motors Work

In its most common mode of operation:

1. An electronic controller sets current output to the motor windings
2. The current defines the position of the internally generated magnetic field
3. A command pulse changes the motor currents to reposition the rotor
4. A series of pulses with a direction command causes rotation in uniformly spaced steps

Speed and Oscillation Behavior

• At sufficiently high pulse frequency, the rotor rotates relatively uniformly but with a modulated velocity
• At the other extreme, a single pulse produces a step followed by an overshoot and a decaying oscillation
• Where the application cannot permit oscillation, damping can be included in the controller

Why Stepper Motors Are Preferred

Stepper motors are often chosen because rotor position is known from the number of pulses and the step size. An initial index point is required as an output position reference.

Critical caution: Care is required in the electronic circuits to avoid introducing random pulses that will cause false positions. As a minimum, the output index point on an appropriate shaft can verify the step count during operation.

Gearing in Motion Control: More Than Just Speed Reduction

In a closed-loop system, gearing couples a high-speed, low-torque motor to a lower-speed, higher-torque load. The gearing must meet requirements for:

• Accuracy
• Strength
• Reliability

Backlash at the Feedback Sensor

The closed loop requires minimum backlash at the point where the feedback sensor is coupled.

For velocity-controlled systems: The feedback sensor is a tachometer usually coupled directly to the rotor shaft. Backlash between motor and tachometer — as well as torsional compliance — must be minimized. Units combining motor and tachometer on a single shaft can usually be purchased as an assembly.

For positioning systems: A position feedback sensor is closely coupled to the shaft being positioned. Antibacklash gearing is frequently used between the gearing and the position feedback sensor. When the position feedback sensor is a limited-rotation device, it may be coupled to a gear that turns faster than the output gear to allow use of its full range.

The Stiffness Requirement

When an appreciable load inertia is being driven, it is critical that the mechanical stiffness between the position sensor coupling point and the load be high enough to avoid natural torsional resonances in the passband.

Feedback Transducers: The Eyes and Ears of the Control System

This is where the practitioner's story began — and where many motion control failures originate. Feedback transducers are the key to accuracy in closed-loop systems.

When the accuracy of a carefully designed control system approaches the accuracy of the feedback transducer, the need for precision in the other system components is reduced. The transducer becomes the accuracy bottleneck — and the accuracy ceiling.

What Transducers Measure

Transducers may measure the quantity being controlled in digital or analog form, and are available for many parameters:

• Pressure and temperature
• Distance traveled or degrees of rotation
• Linear or rotary motion
• Velocity, position, and sometimes acceleration

Although some transducers are nonlinear, a linear relationship between the measured variable and the (usually electrical) output is most common.

Understanding Transducer Errors

The output characteristics of an analog linear-position transducer contain discrete error components that can be distinguished — and individually canceled — by external adjustments:

The most common compensation is for zero-position alignment. The second most common is output gradient adjustment (volts per degree).

Computer-Enhanced Accuracy

Greater accuracy can be achieved in computer-controlled systems by using the computer to cancel out transducer errors. The system's mechanical values and corresponding transducer values are stored in a lookup table and referred to as necessary. Accuracies approaching the inherent repeatability and stability of the system can thus be secured — with recalibration performed at frequent intervals.

Analog Transducers: From Potentiometers to Differential Transformers

Resistance Potentiometers

The simplest analog position transducer is the resistance potentiometer, using a deposited-film rather than wire-wound resistance element.

Performance specifications of modern conductive-plastics potentiometers:

Limitation: Performance deteriorates at high speeds. Prolonged operation above 10 rpm causes excessive wear and increasing output noise.

Variable Differential Transformers (VDTs)

An alternative to the resistance potentiometer, the VDT uses electrical coupling between AC magnetic elements to measure angular or linear motion without sliding contacts.

Advantages:

• Unlimited resolution
• Accuracy comparable to the best resistance potentiometers

Requirements:

• AC energization source
• Precision demodulator (to change AC output to DC)
• Phase-angle matching and accurate amplitude-scale factor
• Temperature compensation (primarily due to changes in resistance of copper windings)

Transducer manufacturers will supply full sets of compatible electronic controls.

Synchros and Resolvers: The Workhorse of Precision Angle Measurement

Synchros and resolvers are transducers widely used for sensing angles at accuracies down to 10 to 20 arc-seconds. More typically (and at much lower cost), their accuracies are 1 to 2 arc-minutes. Cost is further reduced when accuracies of 0.1 degree or higher are acceptable.

Synchro Types and Characteristics