Author: Nima Rad
Communication Protocol
Table of Contents
- General Structure of Controller-to-Axis Communication
- Pulse-Based Control Methods
- Types of Network Communication Protocols
- Manufacturer-Specific Networks
- The Difference Between Standard Ethernet and Real-Time Ethernet
- Choosing the Right Communication Protocol for a CNC
- The Advantages of Digital Networks in Motion Control
- Important Tips for Choosing a Protocol, Installation, and Cabling
- Well-Known Brands and Ecosystems
- Common Mistakes in Choosing a Communication Method
- Summary
Introduction
The communication protocol between a controller and motion axes is one of the most important parts of designing any CNC system, industrial robot, or motion control system. In these machines, the controller has to send motion commands to the servo drive or axis drive, and on many systems, it also needs to receive information such as the axis’s actual position, speed, torque, servo drive status, and its faults.
How the controller communicates with the servo drive has a direct effect on motion accuracy, response speed, coordination between axes, machining quality, the number of controllable axes, the amount of wiring, and the system’s troubleshooting capability.
On older, simpler systems, motion commands are usually transmitted as pulses or an analog signal. The most important pulse-based methods are:
- Pulse/Direction
- CW/CCW
- AB-Phase or Quadrature Pulse
Some older industrial systems also use a ±10V analog command.
On the newer generation of CNC controllers, using real-time digital networks such as CANopen, EtherCAT, PROFINET IRT, EtherNet/IP CIP Motion, Sercos, and MECHATROLINK has become far more common.
An important point is that methods such as Pulse/Direction, CW/CCW, AB-Phase, and ±10V aren’t technically considered “network protocols” — but since they’re among the most important ways of communicating between a controller and a drive, this article covers them alongside network communication protocols.
General Structure of Controller-to-Axis Communication
A simple servo axis setup usually looks like this:
CNC/Motion Controller ← Command/Communication Interface ← Servo Drive ← Servo Motor → Encoder
On most newer servo systems, the encoder connects directly to the servo drive.
The servo drive runs fast internal loops such as the current loop, the velocity loop, and in some cases the position loop.
The CNC controller, for its part, handles tasks such as calculating the motion path, interpolation, coordinating multiple axes, and generating the position setpoint, velocity setpoint, and torque setpoint. So how information is exchanged between the controller and the drive is one of the most important parts of designing a motion control system.
Pulse-Based Axis Control Methods
One of the oldest and most common ways to control a servo drive or stepper drive is sending commands as pulses. In this method, the axis’s travel is usually determined by the number of pulses, and its speed by the pulse frequency.
Different servo drives may support one or more of the following methods for receiving pulse commands: Pulse/Direction, CW/CCW, and AB-Phase or Quadrature Pulse.
The Pulse/Direction Method
This is one of the most common types of pulse command in CNC.
There are two main signals: Pulse and Direction. The Pulse signal sets the number and speed of the pulses, and the Direction signal sets the direction of motion.
For example, if the servo drive’s electronic gear is set so that 10,000 pulses equal one motor revolution, sending 10,000 pulses will cause the motor to turn once. If the pulse frequency increases, the motion speed increases too. Direction sets whether the motion happens in the positive or negative direction.
Advantages of Pulse/Direction
- Very simple structure
- Relatively easy setup
- Low cost
- Supported by many servo drives
- Widely supported by stepper drives
- Not heavily tied to one brand
- Suitable for simple to medium CNCs
Disadvantages of Pulse/Direction
- Needs separate wiring for each axis
- Limited maximum pulse frequency
- Wiring increases as axis count grows
- Very limited feedback information
- Weaker diagnostics than digital networks
- Harder to achieve precise multi-axis synchronization compared to real-time networks
Pulse/Direction is still one of the most common control methods for economical and mid-range CNC machines.
The CW/CCW Method
Another way to send a pulse command is CW/CCW. CW stands for Clockwise and CCW for Counter-Clockwise.
This method uses two independent pulse paths: CW Pulse and CCW Pulse. Sending a pulse to the CW input moves the axis in one direction, and sending a pulse to the CCW input moves it in the opposite direction. So unlike Pulse/Direction, there’s no separate direction signal — the line the pulse is sent on determines the direction of motion.
Advantages of CW/CCW
- Simple structure
- Sets direction without a separate Direction line
- Compatible with a large number of servo drives
- Suitable for some older, industrial systems
Disadvantages of CW/CCW
- Needs two pulse channels
- Limited pulse frequency
- Independent wiring for each axis
- Limited diagnostics
- Less flexible than newer motion networks
The AB-Phase or Quadrature Pulse Method
A third pulse control method is AB-Phase, or Quadrature Pulse. In this method, the controller generates two pulse trains called Phase A and Phase B. These two signals have a 90-degree electrical phase difference. The direction of motion is determined by whether Phase A leads or lags Phase B.
For example, in one direction, A may change state before B, and in the opposite direction, B changes state before A. So in this method, the direction of motion is detected from the phase relationship between the two signals, and there’s no separate Direction line.
How direction is determined in AB-Phase
When moving in one direction: A leads B (meaning Phase A is ahead of Phase B). In the opposite direction: B leads A (meaning Phase B is ahead of Phase A). The drive detects the direction of motion by checking the order of the A and B edges.
AB-Phase resolution
One important feature of a quadrature signal is that the receiver can count different signal transitions depending on the hardware design. Counting may be done as x1, x2, or x4. In x4 mode, all four edges related to A and B can be used in the count. So at a given signal frequency, a higher counting resolution can be achieved. That said, this depends on the servo drive’s input decoding method, and the drive’s specifications need to be checked.
The difference between an AB-Phase command and encoder A/B
In terms of waveform, an AB-Phase command looks a lot like an incremental encoder’s output. Incremental encoders also usually have A, B, and Z signals, but their use is different. In an AB-Phase command, the controller sends the A/B signal to the drive to generate a motion command. With an encoder’s A/B output, the encoder sends the A/B signal to the controller or drive so the axis’s actual position can be measured. So the signal shape is similar, but the direction of information transfer and its use are different.
Advantages of AB-Phase
- Can determine direction without a Direction line
- Standard, well-known quadrature structure
- Can achieve higher resolution with x2 or x4 decoding
- Compatible with some industrial servo drives
- Suitable for some older controllers and motion systems
Disadvantages of AB-Phase
- Requires precisely generating two pulses with a defined phase difference
- More hardware complexity than Pulse/Direction
- Limited by the drive’s input frequency
- Still has the general limitations of pulse control
- Gives the controller limited feedback and diagnostic information
On newer systems, AB-Phase, like other pulse command methods, has gradually been replaced by real-time networks in many advanced applications.
Comparing the Three Pulse Control Methods
Pulse/Direction: Signals are Pulse + Direction; pulse count sets the travel distance and pulse frequency sets the speed; Direction sets the direction; low complexity; very common in CNC and stepper/servo systems.
CW/CCW: Signals are CW Pulse + CCW Pulse; CW commands motion in one direction and CCW in the opposite direction; low complexity; used in servo drives and various industrial systems.
AB-Phase: Signals are Phase A + Phase B with a 90-degree phase difference; direction is detected from the lead/lag between A and B; medium complexity; used in some servo systems, motion controllers, and quadrature-signal-based systems.
±10V Analog Control
Another traditional way to control a servo drive is an analog command. In this method, the controller usually generates a voltage in the range of -10V to +10V for the servo drive; for example, +10V equals maximum speed in the positive direction, 0V equals zero speed, and -10V equals maximum speed in the negative direction.
On many older industrial CNCs, the position loop ran inside the CNC controller, and based on the position error, the controller sent a velocity command as ±10V to the servo drive.
Advantages: Fast response, well-known industrial architecture, suitable for older CNCs, allows running the position loop in the CNC controller.
Disadvantages: Sensitive to noise, possible offset, possible drift, needs a precise analog output, needs suitable cabling, limited diagnostics, more cabling.
Today, on newer systems, real-time digital networks have replaced this method in many applications.
Types of Network Communication Protocols
CANopen
CANopen is an industrial communication protocol based on CAN bus. It’s used in automation, servo control, positioning systems, and mobile machines.
One of the most important CANopen standards for servo drives is CiA 402. CiA 402 provides a standardized structure for controlling drives. Parameters such as Controlword, Statusword, Target Position, Actual Position, Target Velocity, Actual Velocity, and Target Torque are defined within this structure.
Advantages: Well-known industrial standard, simple wiring, can connect several drives on one bus, better diagnostics than pulse, can read drive status, can adjust parameters from the controller.
Disadvantages: Narrower bandwidth than industrial Ethernet, limited on very fast systems, weaker synchronization than EtherCAT, not suitable for some very high-performance applications.
EtherCAT
EtherCAT is one of the most important real-time Ethernet communication protocols in modern motion control.
A system’s structure might look like: Motion Controller ← EtherCAT ← Servo Drive X ← Servo Drive Y ← Servo Drive Z ← Servo Drive A ← Remote I/O. In this architecture, one network cable can connect a large number of devices to one another.
Distributed Clocks
One of EtherCAT’s most important capabilities is Distributed Clocks (DC). Distributed Clocks allow very precise synchronization of network nodes. This matters a great deal for things like multi-axis interpolation, gantry control, electronic cam, robotics, and CNC machining.
EtherCAT and CiA 402
Many EtherCAT servo drives use CoE (CANopen over EtherCAT). In this case, the CiA 402 structure is made available to the controller through EtherCAT. Among the most important motion modes are:
- CSP (Cyclic Synchronous Position): The controller sends a new target position value to the drive every cycle. This mode is very useful for CNC and robotics.
- CSV (Cyclic Synchronous Velocity): The controller sends a target velocity value to the servo drive every cycle.
- CST (Cyclic Synchronous Torque): The controller sends a target torque value. This mode is used in applications such as force control, robotics, tension control, and advanced motion.
Advantages of EtherCAT: Low cycle time, high speed, very precise synchronization, suitable for large axis counts, simpler cabling, powerful diagnostics, can transfer position, velocity, and torque, can read drive parameters, suitable for CNC and robotics, suitable for high-performance motion.
Disadvantages of EtherCAT: More complex master design than pulse, needs more technical knowledge, the controller’s real-time performance matters a great deal, needs correct cycle-time design, needs correct synchronization settings.
PROFINET
PROFINET is an important industrial Ethernet communication protocol. In motion control, a distinction needs to be made between PROFINET RT and PROFINET IRT. IRT stands for Isochronous Real-Time and is designed for applications that need high synchronization. PROFIdrive is typically used to control drives.
Advantages: Broad industrial ecosystem, good integration with PLCs, strong diagnostics, suitable for production lines, real-time capability, suitable for multi-axis motion.
Disadvantages: More complex design, higher cost for some equipment, needs precise network configuration.
PROFINET is especially common in the Siemens ecosystem.
EtherNet/IP and CIP Motion
EtherNet/IP is another industrial Ethernet communication protocol. CIP Motion is used for motion control, and CIP Sync is used for time synchronization. CIP Motion is used to transfer position, velocity, and torque information and to control multiple axes.
Advantages: Large industrial ecosystem, integration capability with PLCs, suitable diagnostics, suitable for automation, advanced motion control capability.
Disadvantages: More complex than pulse-based methods, needs correct network design, performance depends on the system’s architecture.
This technology is very widely used in the Rockwell Automation ecosystem.
Sercos
Sercos stands for Serial Real-time Communication System. This communication protocol was designed from the start for motion control and servo drives. Its newer version, Sercos III, uses Ethernet.
Advantages: Precise synchronization, dedicated motion design, suitable for multi-axis systems, suitable real-time performance, standardized drive parameters.
Disadvantages: Smaller ecosystem than EtherCAT and PROFINET, less equipment available in some markets, less general technical familiarity.
MECHATROLINK
MECHATROLINK is a specialized motion control communication protocol. This technology is particularly associated with Yaskawa servo systems. Its versions include MECHATROLINK-II, MECHATROLINK-III, and MECHATROLINK-4.
Advantages: Suitable servo motion performance, good synchronization, suitable for multi-axis, good drive-and-controller integration.
Disadvantages: More dependent on a specific ecosystem, less widespread than EtherCAT.
POWERLINK
Ethernet POWERLINK is also a real-time industrial Ethernet communication protocol. It’s used for real-time communication, synchronization, automation, and multi-axis motion. POWERLINK has been used on some B&R systems and other industrial equipment.
Modbus RTU and Modbus TCP
Modbus is a common industrial communication protocol with two well-known versions: Modbus RTU and Modbus TCP. Many servo drives and inverters support Modbus. But Modbus usually isn’t a suitable choice for fast, coordinated multi-axis motion.
Its suitable applications include Start/Stop, reading alarms, reading status, parameter setting, adjusting inverter speed, monitoring, and diagnostics.
So having Modbus on a servo drive doesn’t necessarily mean it’s suitable for fast CNC motion.
Manufacturer-Specific Networks
Some manufacturers offer their own proprietary networks, including SSCNET, RTEX, MECHATROLINK, and other dedicated networks.
The main advantage of such networks is very good integration between one manufacturer’s equipment. But one of their disadvantages can be vendor lock-in — meaning the controller and drive become more dependent on one company’s ecosystem.
The Difference Between Standard Ethernet and Real-Time Ethernet
Having an RJ45 port doesn’t mean a network is real-time. Standard Ethernet is suitable for things like HMI, SCADA, programming, monitoring, and file transfer.
But controlling servo axes requires determinism. Determinism means data has to be transferred within a defined, predictable time window.
Another very important parameter is jitter. Jitter is how much the execution time of consecutive cycles varies. The lower the jitter, the more precisely different axes can be synchronized.
The Importance of Cycle Time
Cycle time determines how often the controller sends new motion information to the drive. For example, a cycle time of 1ms means the controller updates motion information roughly 1,000 times per second.
On some advanced systems, cycle time can be even lower, but lowering cycle time doesn’t always mean better quality. The controller’s CPU, real-time operating system, network master, servo drive, motion algorithm, and the number of network nodes all need to be able to run the intended cycle; otherwise, jitter or communication errors may increase.
Centralized Motion Control
In a centralized motion control architecture, the main motion calculations happen inside the controller. The controller can send the target position, target velocity, and target torque to the drives every cycle. This architecture is very well suited to multi-axis CNCs with precise interpolation.
Distributed Motion Control
In distributed motion control, part of the motion calculation happens inside the servo drive or motion node. In this system, the controller sends higher-level commands. Choosing between a centralized and distributed architecture depends on the type of machine and application.
Choosing the Right Communication Protocol for a CNC
For a simple 3- or 4-axis CNC where cost and simplicity matter more, pulse-based methods such as Pulse/Direction, CW/CCW, and AB-Phase can be suitable options. Among these, Pulse/Direction is one of the most common choices today.
But on machines with more axes, higher speed, gantry axes, advanced interpolation, a need for precise synchronization, robotics, or electronic cam, real-time networks offer a much bigger advantage — including EtherCAT, PROFINET IRT, Sercos, and CIP Motion.
The Advantage of a Digital Network Isn’t Just Speed
One common mistake is assuming that EtherCAT and other motion networks’ only advantage is higher speed, when a digital network actually has other important advantages too:
Reduced cabling In a pulse system, each axis needs its own independent command lines. On digital networks, several drives can share one common bus.
Diagnostics The controller can receive information such as drive errors, status word, actual position, actual velocity, torque, following error, and temperature.
Parameter management Many drive parameters can be read and adjusted directly over the network.
Synchronization All servo axes can work in coordination based on one shared time reference.
Important Tips for Choosing a Communication Protocol
The following should be checked to choose a suitable system:
- Number of axes on the machine
- Type of servo drive
- Required speed
- Machine accuracy
- Resolution
- Cycle time
- Jitter
- Type of interpolation
- Electronic gear
- Electronic cam
- Gantry synchronization
- Diagnostics
- Safety
- Number of I/O points
- System cost
- Availability of equipment
- Ability to use different brands
- Room for future machine development
- The maintenance team’s technical knowledge
Important Installation and Cabling Tips
Communication quality doesn’t depend on the protocol alone. Poor cabling can cause problems even on a very fast network.
Important points include:
- Using standard cable
- Suitable shielding
- Correct grounding
- Keeping the network cable separate from motor power cable
- Distance from the servo drive and inverter output
- Using a suitable connector
- Respecting the cable’s bend radius
- Preventing ground loops
- Following EMC principles
On a CNC machine, because of the servo drive, spindle drive, and inverter, following these points matters a great deal.
Well-Known Brands and Ecosystems
Different motion control companies use different networks:
Beckhoff One of the most important companies associated with EtherCAT.
Siemens Has a broad ecosystem built around PROFINET and PROFIdrive.
Rockwell Automation Makes extensive use of EtherNet/IP and CIP Motion.
Bosch Rexroth Has a long track record in motion control and industrial Ethernet.
Yaskawa Known for servo motion and MECHATROLINK.
Omron Uses EtherCAT on many of its newer motion controllers.
Mitsubishi Electric Has several motion ecosystems and industrial networks.
Panasonic Offers servo drives with pulse interfaces and various digital networks.
A servo drive shouldn’t be chosen based on its communication protocol alone; factors such as servo tuning, encoder resolution, dynamic response, drive power, motor quality, safety features, service, and parts availability also need to be checked.
Common Mistakes in Choosing a Communication Method
Choosing a network based only on Mbps High bandwidth alone doesn’t indicate motion quality. Parameters such as determinism, cycle time, synchronization, and jitter matter more.
Treating standard Ethernet and real-time Ethernet as the same thing Not every network that uses Ethernet is suitable for servo motion.
Choosing a drive based only on having EtherCAT Having an EtherCAT port isn’t enough. It needs to be checked which modes the drive supports, such as CSP, CSV, and CST.
Using Modbus for fast motion Modbus wasn’t designed for fast, coordinated control of multiple servo axes.
Reducing cycle time too aggressively A low cycle time only helps if every part of the system can run it reliably.
Confusing an AB-Phase command with encoder feedback A/B signals can be used both as a command and as an encoder output. So when designing the system, it needs to be clear whether the A/B signal in question is a command input or encoder feedback.
Summary
The communication protocol between a CNC controller and a servo drive is one of the main factors determining a motion control system’s performance.
On simpler systems, pulse-based methods such as Pulse/Direction, CW/CCW, and AB-Phase are still widely used. Pulse/Direction, thanks to its simplicity and broad compatibility, is one of the most common methods. CW/CCW sets the direction of motion using two independent pulse channels. In the AB-Phase method, two signals, A and B, are generated with a 90-degree phase difference, and the direction of motion is detected from whether one signal leads or lags the other. A ±10V analog command is also still seen on older industrial CNCs and some retrofits.
As axis count, machine speed, and the need for precise diagnostics and synchronization increase, using real-time networks becomes far more important. CANopen enables networked control of drives, while technologies such as EtherCAT, PROFINET IRT, EtherNet/IP CIP Motion, Sercos, and MECHATROLINK are designed for more advanced motion control.
Among them, EtherCAT — thanks to its low cycle time, Distributed Clocks, multi-axis capability, good diagnostics, and broad servo drive support — is one of the important options for CNC, robotics, and modern motion control systems.
That said, no single communication protocol can be considered the best choice for every machine. On a simple CNC, Pulse/Direction may be entirely sufficient, while a 5-axis CNC, a gantry machine, a robot, or a fast multi-axis system may need a real-time network such as EtherCAT.
So the right choice should be based on the machine’s actual needs, axis count, accuracy, speed, cycle time, motion architecture, equipment cost, and room for future system development.
For product evaluation, compare Radonix PC-based industrial CNC controllers, expansion hardware, software and machine-specific interfaces. Radonix industrial CNC controller range.
