ELMO Academy


Achieving precise, coordinated movement across multiple servo axes demands deterministic communication and tight timing. EtherCAT multi-axis synchronization delivers exactly that, enabling machines to move with sub-microsecond accuracy. This guide explores distributed clock alignment, coordinated torque feedforward, one-cable technology, and practical strategies for cycle time reduction in demanding applications like CNC lathe motion control.

Key Takeaways

  • Distributed clocks are non-negotiable: Hardware-level time alignment is the prerequisite for all forms of multi-axis synchronization. Software-based approaches cannot match the consistency that EtherCAT DC provides.
  • Jitter determines motion quality: Sub-microsecond jitter translates directly into lower contour errors, better surface finishes, and more consistent part quality.
  • Feedforward and current-loop tuning multiply the benefit: Synchronization accuracy enables advanced control strategies that would be ineffective on less deterministic networks.
  • Practical benefits are measurable: Cycle time reduction, improved part quality, and simplified wiring through one-cable technology deliver quantifiable returns on investment.

What Is EtherCAT Multi-Axis Synchronization?

EtherCAT multi-axis synchronization refers to the ability of multiple servo drives, connected over an EtherCAT fieldbus network, to execute motion commands at precisely the same instant. Unlike traditional fieldbus architectures that process commands sequentially with variable latency, EtherCAT uses a unique “processing on the fly” telegram structure. Each node reads and inserts data into a single Ethernet frame as it passes through, eliminating the store-and-forward delays that plague conventional networks.

Why Synchronization Matters in Multi-Axis Systems

When a machine requires two or more axes to trace a coordinated path, such as an interpolated contour on a milling machine or a synchronized pick-and-place trajectory, even microseconds of timing mismatch between drives can produce visible surface defects, positional errors, or mechanical stress. Synchronization ensures that every axis receives its setpoint and begins execution within the same narrow time window.

Key Characteristics of EtherCAT-Based Synchronization

  • Deterministic telegram processing: Each drive processes its portion of the EtherCAT frame in hardware, with propagation delays measured in nanoseconds per node.
  • Distributed clock mechanism: A shared time reference aligns all nodes to a common clock, enabling simultaneous actuation regardless of physical position on the network.
  • High bandwidth: EtherCAT supports update rates exceeding 10 kHz for dozens of axes, making it suitable for both high-speed and high-precision applications.
  • Scalability: Adding axes to an EtherCAT network does not degrade synchronization quality, because the distributed clock protocol compensates for propagation delays automatically.

Manufacturers like Elmo have built their servo drive platforms around EtherCAT precisely because of these synchronization capabilities, offering compact drives that maintain deterministic performance even in densely packed multi-axis configurations.

The Core Principle: How Distributed Clock Alignment Works

Distributed clock alignment is the foundation that makes EtherCAT multi-axis synchronization possible. Without a shared, high-resolution time base, individual drives would each rely on their own local oscillators, which inevitably drift relative to one another. The EtherCAT distributed clock (DC) protocol solves this problem at the hardware level.

The Reference Clock and Propagation Delay Measurement

One node on the EtherCAT network, typically the first slave device, is designated as the reference clock. All other nodes synchronize their local clocks to this reference. During network initialization, the EtherCAT master sends special datagrams that allow each node to measure the propagation delay between itself and the reference clock. These measurements account for cable lengths, connector delays, and processing latency within each node’s EtherCAT Slave Controller (ESC).

Continuous Clock Correction

After the initial delay measurement, each slave continuously adjusts its local clock to stay aligned with the reference. This correction happens through a control loop embedded in the ESC hardware:

  1. The master periodically writes its system time to the reference clock node.
  2. Each slave compares its local time to the propagation-delay-compensated reference time.
  3. A drift compensation algorithm adjusts the local oscillator frequency to minimize the offset.
  4. The result is clock alignment across all nodes, typically within tens of nanoseconds.

SYNC0 and SYNC1 Signals

Once clocks are aligned, the EtherCAT DC mechanism generates hardware interrupt signals called SYNC0 and SYNC1. These interrupts fire at precisely the same absolute time on every node, triggering the drive firmware to latch inputs, compute control loops, and apply outputs simultaneously. SYNC0 typically triggers the main control cycle, while SYNC1 can be used for a secondary event, such as staggering ADC sampling from PWM switching to reduce noise.

This hardware-level time alignment is what separates EtherCAT from software-synchronized fieldbuses and is the reason distributed clock alignment delivers the consistency required for precision multi-axis motion.

Achieving Ultra-Precise Timing with Sub-Microsecond Jitter

Synchronization accuracy is only as good as the timing consistency between cycles. Jitter, the variation in the actual execution time of a control event from one cycle to the next, directly impacts trajectory accuracy. EtherCAT networks routinely achieve sub-microsecond jitter, and well-designed drive systems push this figure below 100 nanoseconds.

Sources of Jitter and How EtherCAT Minimizes Them

Jitter Source Typical Impact EtherCAT Mitigation
Network stack processing 10-100 µs on standard Ethernet Hardware-based frame processing in ESC bypasses the software stack entirely
Operating system scheduling 1-50 µs on non-RT OS SYNC0 interrupt is hardware-generated, independent of master OS scheduling
Clock drift between nodes 1-10 µs without correction Continuous distributed clock correction keeps drift below 100 ns
Telegram arrival variation Variable with network load Deterministic frame structure ensures fixed arrival times

Why Sub-Microsecond Jitter Matters for Motion Quality

Consider a servo axis running at 4,000 RPM with a 125 µs control cycle. A 1 µs jitter corresponds to a positional uncertainty of approximately 0.024 degrees per cycle. For a single axis this may be tolerable, but when multiple axes must interpolate along a complex path, the cumulative effect of uncorrelated jitter across axes creates contour errors. Reducing jitter to below 1 µs keeps these errors well within the mechanical resolution of most servo systems.

Elmo’s Approach to Minimizing Jitter

Elmo servo drives implement their current and position control loops in dedicated hardware and FPGA logic, ensuring that the time between the SYNC0 interrupt and the actual PWM update remains constant cycle to cycle. This deterministic internal architecture, combined with EtherCAT’s distributed clock, enables sub-microsecond jitter performance even at cycle times as short as 62.5 µs.

The practical result is smoother velocity profiles, lower following errors, and improved surface finish quality in machining applications where timing precision directly translates to part quality.

Advanced Control via Coordinated Torque Feedforward

Precise position tracking at high speeds requires more than just a well-tuned position loop. Coordinated torque feedforward allows the motion controller to anticipate the torque demands of a trajectory and inject the appropriate current commands proactively, rather than waiting for position errors to accumulate and relying solely on feedback correction.

How Torque Feedforward Works in a Multi-Axis Context

In a single-axis system, torque feedforward is straightforward: the controller calculates the expected inertial load from the commanded acceleration profile and adds a proportional torque command to the output of the position loop. In a multi-axis system, the challenge increases because:

  • Axes are mechanically coupled: In robotic arms, gantry systems, and CNC machines, motion on one axis creates reaction forces on others.
  • Gravitational and friction terms vary with pose: The feedforward model must account for the current configuration of the machine.
  • Timing must be exact: Feedforward torque commands are only effective if they arrive at each drive at precisely the right instant relative to the trajectory segment.

The Role of EtherCAT Synchronization in Feedforward Accuracy

Coordinated torque feedforward depends entirely on the synchronization quality of the network. If one drive receives its feedforward command 5 µs late relative to another, the pre-computed torque no longer matches the actual dynamic state of the machine. EtherCAT’s distributed clock ensures that all feedforward values are applied within the same SYNC0 window, preserving the validity of the dynamic model.

Implementation Considerations

  1. Model fidelity: The feedforward model should include inertia, viscous friction, Coulomb friction, and gravity terms at a minimum. For articulated mechanisms, a full inverse dynamics model is preferred.
  2. Trajectory planning: Jerk-limited (S-curve) profiles produce smoother acceleration demands, which improve feedforward effectiveness.
  3. Drive-level support: The servo drive must accept torque feedforward as an additional input alongside the position or velocity command. Elmo drives support this through standard CiA 402 profile modes and custom PDO mappings.
  4. Tuning methodology: Start with acceleration feedforward (proportional to commanded acceleration), then add velocity feedforward if residual tracking error persists at constant speed.

When properly implemented, coordinated torque feedforward can reduce tracking errors by 50-90% compared to pure feedback control, enabling higher contouring speeds without sacrificing accuracy.

Optimizing Dynamics with Shared Current-Loop Optimization

The current loop is the innermost and fastest control loop in a servo drive, typically executing at rates between 10 kHz and 100 kHz. Its performance determines how quickly and accurately the drive can produce the torque commanded by the outer velocity and position loops. Shared current-loop optimization refers to the practice of tuning and coordinating the current loops across all axes in a multi-axis system to achieve consistent, optimal dynamic response.

Why Current-Loop Consistency Matters

If two axes in a gantry system have different current-loop bandwidths, one axis will respond to torque commands faster than the other. This mismatch creates transient contour errors during acceleration and deceleration phases, even if the position loops are identically tuned. Shared current-loop optimization eliminates this asymmetry.

Key Parameters for Optimization

  • Current-loop bandwidth: Should be matched across all coordinated axes. A typical target is 2-4 kHz for standard servo applications and up to 10 kHz for high-performance systems.
  • Phase margin: Each axis should maintain at least 45 degrees of phase margin to prevent oscillation while maximizing responsiveness.
  • PWM frequency: Higher PWM frequencies allow higher current-loop bandwidths but increase switching losses. The trade-off must be balanced against thermal constraints.
  • Sampling synchronization: Current measurements on all axes should be sampled at the same point in the PWM cycle to ensure consistent phase relationships.

Automated Tuning and Commissioning

Manual tuning of current loops across many axes is time-consuming and error-prone. Modern servo drives offer auto-tuning functions that measure motor parameters (resistance, inductance, back-EMF constant) and automatically set current-loop gains. Elmo’s commissioning tools, for example, can identify motor parameters and configure optimal current-loop settings in seconds, then replicate those settings across matched axes to ensure uniform dynamic response.

The result of shared current-loop optimization is a multi-axis system where every axis responds identically to torque commands, providing the consistent foundation that higher-level position and velocity loops require for accurate coordinated motion.

Application Spotlight: Precision CNC Lathe Motion Control

CNC lathe motion control presents a demanding test case for EtherCAT multi-axis synchronization. A modern CNC lathe may require tight coordination between the spindle drive, one or more linear axes (X, Z), a live tooling axis (C or Y), and auxiliary systems like tool turrets and tailstocks. Each of these axes must work in concert to produce parts with tolerances measured in single-digit micrometers.

Synchronization Requirements in Turning Operations

Operation Axes Involved Synchronization Requirement
Constant surface speed turning Spindle + X axis Spindle speed must adjust as X-axis position changes to maintain constant cutting velocity
Thread cutting Spindle + Z axis Z-axis feed must be phase-locked to spindle rotation with zero cumulative error
Live tooling (milling on lathe) C axis + X/Y/Z Interpolated path must be synchronized with C-axis rotary position
Polygon turning Spindle + polygon tool spindle Two spindles must maintain a precise, non-integer speed ratio

How EtherCAT Addresses CNC Lathe Challenges

Thread cutting is perhaps the most demanding synchronization task on a lathe. The Z-axis must advance by exactly one thread pitch per spindle revolution, and any deviation accumulates as a pitch error on the workpiece. With EtherCAT’s distributed clock providing sub-microsecond jitter, the electronic gearing between spindle encoder feedback and Z-axis position command remains phase-accurate across thousands of consecutive revolutions.

Real-World Performance Gains

CNC lathe builders who have migrated from legacy analog or SERCOS-based architectures to EtherCAT with high-performance drives report measurable improvements:

  • Thread pitch accuracy: Improved from ±5 µm to ±1 µm through tighter spindle-to-axis synchronization.
  • Surface finish (Ra): Reduced from 0.8 µm to 0.4 µm on finish turning passes due to lower velocity ripple.
  • Cycle time reduction: 10-15% shorter part cycles through faster acceleration profiles enabled by coordinated torque feedforward.
  • Setup time: Reduced by 30% through standardized EtherCAT configuration and auto-tuning.

Elmo’s compact servo drives are particularly well-suited for CNC lathe applications where cabinet space is limited and high current-loop bandwidth is essential for spindle control responsiveness.

Simplifying Machine Design with EtherCAT One-Cable Technology

Wiring complexity is a significant cost driver in multi-axis machines. Traditional servo installations require separate cables for motor power and encoder feedback for each axis, plus additional wiring for brake control, temperature sensors, and other auxiliary signals. EtherCAT one-cable technology (OCT) consolidates motor power and feedback into a single hybrid cable, dramatically reducing cable count, connector count, and installation labor.

How One-Cable Technology Works

OCT combines the motor’s power conductors with a digital feedback channel in a single cable assembly. The feedback data, which includes position, velocity, temperature, and diagnostic information, is transmitted digitally over dedicated signal wires within the same cable sheath. The digital protocol replaces traditional analog encoder interfaces (sin/cos, resolver) with a high-speed serial link that offers higher resolution and built-in error detection.

Benefits for Multi-Axis Machines

  • 50% fewer cables: Each axis requires one cable instead of two, cutting cable tray volume and reducing drag chain size.
  • Fewer failure points: Fewer connectors mean fewer potential sources of intermittent faults and downtime.
  • Smaller cabinet footprint: Reduced terminal count frees space for other components or allows smaller enclosures.
  • Faster commissioning: Single-cable connections reduce wiring time per axis from approximately 30 minutes to under 10 minutes.

Compatibility with EtherCAT Synchronization

One-cable technology operates independently of the EtherCAT fieldbus layer. The feedback data transmitted over the hybrid cable is processed locally by the drive, while EtherCAT handles the real-time communication between the drive and the motion controller. This separation means that OCT does not introduce any additional jitter or latency into the synchronization chain. The distributed clock alignment and sub-microsecond jitter performance remain identical whether the drive uses a traditional two-cable setup or a single hybrid cable.

Elmo supports one-cable technology across its drive portfolio, allowing machine builders to benefit from simplified wiring without compromising on motion performance or synchronization quality.

The Business Case: Achieving Measurable Cycle Time Reduction

Engineering performance specifications like jitter and synchronization accuracy ultimately need to translate into business outcomes. Cycle time reduction is the most direct and measurable benefit of improved EtherCAT multi-axis synchronization, and it impacts throughput, cost per part, and return on capital equipment investment.

Where Cycle Time Savings Come From

Faster cycle times in synchronized multi-axis systems are not simply the result of increasing axis speeds. They emerge from several interrelated improvements:

  1. Higher acceleration without tracking error: Coordinated torque feedforward allows more aggressive acceleration profiles while maintaining path accuracy, reducing the time spent ramping up and down.
  2. Shorter settling times: Matched current-loop dynamics and reduced jitter produce faster settling at target positions, eliminating wasted dwell time.
  3. Tighter contouring at speed: Better synchronization allows the machine to maintain higher feedrates through curves and corners without exceeding contour tolerance.
  4. Reduced scrap and rework: Consistent part quality means fewer rejected parts and less time spent on secondary finishing operations.

Quantifying the Impact

Improvement Area Typical Cycle Time Savings Annual Impact (3-shift operation)
Higher acceleration profiles 5-10% per move 200-400 additional production hours
Reduced settling time 3-8% per cycle 120-320 additional production hours
Higher contouring feedrate 5-15% on complex paths 200-600 additional production hours
Reduced scrap rate 1-3% material savings Direct cost reduction

Calculating ROI

For a machine producing 500 parts per day with a 30-second cycle time, a 10% cycle time reduction yields 50 additional parts per day, or approximately 18,000 additional parts per year on a three-shift schedule. If each part contributes $5 in margin, the annual benefit is $90,000, a figure that often exceeds the total cost of the drive and motion control system within the first year of operation.

These calculations make a compelling case for investing in high-performance EtherCAT drive systems that deliver the synchronization accuracy needed to extract maximum productivity from multi-axis machines.

Essential PLCopen Function Blocks for Synchronized Motion

Implementing EtherCAT multi-axis synchronization in a real machine requires standardized software interfaces. The PLCopen Motion Control standard defines a set of function blocks that provide a vendor-independent programming model for single-axis and multi-axis motion. Understanding the key function blocks is essential for engineers designing synchronized motion applications.

Single-Axis Foundation Blocks

  • MC_Power: Enables or disables the drive. Must be called before any motion command.
  • MC_MoveAbsolute / MC_MoveRelative: Commands a single axis to move to an absolute or relative position.
  • MC_MoveVelocity: Commands continuous motion at a specified velocity.
  • MC_Home: Executes a homing sequence to establish a known reference position.
  • MC_Stop: Decelerates the axis to a standstill using a defined deceleration ramp.

Multi-Axis Coordination Blocks

  • MC_CamIn / MC_CamOut: Establishes or releases an electronic cam relationship between a master and slave axis. Used extensively in packaging, printing, and rotary motion applications.
  • MC_GearIn / MC_GearOut: Creates an electronic gear ratio between two axes. Critical for thread cutting on CNC lathes and winding applications.
  • MC_MoveLinearAbsolute / MC_MoveLinearRelative: Commands coordinated linear interpolation across multiple axes simultaneously.
  • MC_MoveCircularAbsolute: Commands circular interpolation in a defined plane, essential for CNC contouring.
  • MC_MovePath: Executes a pre-defined multi-segment path with blending between segments.

Synchronization-Specific Blocks

Beyond basic coordination, several function blocks address the timing and synchronization layer directly:

  1. MC_SyncMasterAxis: Designates a master axis for synchronization groups and manages the phase relationship between master and slave axes.
  2. MC_PhasingAbsolute / MC_PhasingRelative: Adjusts the phase offset between synchronized axes without breaking the gear or cam relationship.
  3. MC_TouchProbe: Captures the exact position of an axis at the moment a digital input triggers, using the distributed clock timestamp for sub-microsecond accuracy.

These standardized function blocks allow motion application code to be portable across different controller platforms while still leveraging the full synchronization capabilities of the underlying EtherCAT network. Elmo drives are fully compatible with the PLCopen motion function block standard, enabling straightforward integration with major PLC and motion controller platforms.

Frequently Asked Questions

How does EtherCAT multi-axis synchronization achieve sub-microsecond timing accuracy?

It uses a hardware-level distributed clock protocol that continuously corrects local oscillator drift across all nodes, keeping clock alignment within tens of nanoseconds and generating simultaneous SYNC0 interrupts.

Why does jitter reduction matter for coordinated torque feedforward?

Feedforward torque commands are pre-computed for a specific instant; even a few microseconds of jitter causes the torque to mismatch the machine’s actual dynamic state, degrading tracking accuracy across synchronized axes.

What cycle time reduction can manufacturers expect from improved EtherCAT multi-axis synchronization?

Combining higher acceleration profiles, shorter settling times, and faster contouring feedrates typically yields 10–15% shorter part cycles, translating to hundreds of additional production hours annually.

How does one-cable technology affect sub-microsecond jitter performance in synchronized systems?

EtherCAT one-cable technology operates independently of the fieldbus layer, so consolidating power and feedback into a single hybrid cable introduces no additional latency or jitter to the distributed clock alignment chain.

What role does shared current-loop optimization play in CNC lathe motion control?

Matching current-loop bandwidth and phase margin across all coordinated axes eliminates transient contour errors during acceleration, ensuring uniform torque response critical for operations like thread cutting and finish turning.

Which PLCopen function blocks are essential for EtherCAT multi-axis synchronization in thread cutting?

MC_GearIn establishes the electronic gear ratio phase-locking the Z-axis feed to spindle rotation, while MC_TouchProbe uses distributed clock timestamps for sub-microsecond position capture accuracy.

How does coordinated torque feedforward contribute to cycle time reduction on multi-axis machines?

It allows more aggressive acceleration profiles without increasing tracking error, reducing ramp-up and ramp-down time per move by 5–10% while maintaining path accuracy across all synchronized axes.

Conclusion: The Future of High-Performance Motion Control

EtherCAT multi-axis synchronization has established itself as the dominant approach for machines that demand precise, coordinated motion across multiple axes. The combination of distributed clock alignment, sub-microsecond jitter, and deterministic communication creates a foundation on which advanced techniques like coordinated torque feedforward and shared current-loop optimization can deliver their full potential.

Where the Technology Is Heading

Several trends are shaping the next generation of synchronized motion systems. EtherCAT G and EtherCAT G10, operating at 1 Gbit/s and 10 Gbit/s respectively, will enable even faster update rates and support for more axes per network. Integration of functional safety over EtherCAT (FSoE) allows safety functions like Safe Torque Off and Safely Limited Speed to operate over the same network without additional wiring. Machine learning-based auto-tuning is beginning to optimize not just individual axes but the interaction dynamics between coupled axes in real time.

For machine builders and system integrators focused on precision CNC lathe motion control, robotics, semiconductor manufacturing, or any application where microseconds matter, investing in a high-quality EtherCAT drive platform is a strategic decision. Companies like Elmo, with deep expertise in compact, high-performance servo drives and a proven track record in demanding multi-axis applications, are well-positioned to support the next generation of precision motion machines.

The path to higher throughput, tighter tolerances, and simpler machine architectures runs through EtherCAT multi-axis synchronization. The technology is mature, the standards are established, and performance gains are real.