summary
Real-time deterministic Ethernet protocols (such as EtherCAT) have enabled the synchronous operation of multi-axis motion control systems. This synchronization has two aspects. First, the transmission of commands and instructions between control nodes must be synchronized with a common clock; second, the execution of control algorithms and feedback functions must also be synchronized with the same clock. The first type of synchronization is easy to understand; it is an inherent part of network controllers. However, the second type of synchronization has been largely overlooked until now and has become a bottleneck for motion control performance.
This paper introduces a novel concept for maintaining motor drive synchronization throughout the entire process, from the network controller to the motor terminals and sensors. The proposed technique significantly improves synchronization, thereby substantially enhancing control performance.
Problem Statement and Prior Art
To illustrate the limitations of existing solutions, consider a two-axis networked motion control system, as shown in Figure 1. The motion control master sends commands and instruction values to two servo controllers via a real-time network, with each servo controller constituting a slave node on the network. Each servo controller itself consists of a network controller, a motor controller, a power inverter, and a motor/encoder.
Real-time network protocols employ various methods to synchronize slave nodes with the master node. A common approach is to configure a local synchronization clock at each node. This consensus on time ensures that command values and instructions for all servo axes are closely synchronized. In other words, all network controllers on a real-time network remain synchronized.
Typically, there are two interrupt lines between the network controller and the motor controller. The first line notifies the motor controller when it needs to collect input and put it on the network. The second line notifies the motor controller when to read data from the network. Following this method, the motion controller and the motor controller exchange data synchronously, achieving very high timing accuracy. However, simply transmitting synchronous data to the motor controller is not enough; the motor controller must also be able to respond to data synchronously. Without this capability, the motor controller cannot fully utilize the timing accuracy of the network. Problems will arise with the motor controller's I/O when responding to command values and commands.
Each I/O in a motor controller (such as a pulse width modulation (PWM) timer and an ADC) has inherent delays and time quantization. For example, consider Figure 2, which shows a PWM timer that generates the gate drive signal for a power inverter. This timer generates the gate signal by comparing the instruction value Mx with an up/down counter. When the control algorithm changes Mx, the new duty cycle doesn't take effect until the next PWM cycle. This is equivalent to a zero-order hold effect, meaning the duty cycle is updated only once or twice per PWM cycle T (if using double-update mode).
Figure 1. Typical two-axis networked motion control system
Figure 2. Duty cycle update of PWM timer
In real-time networks, regardless of how tightly synchronized the data exchange is, the time quantization of the PWM timer ultimately becomes the decisive factor for axis synchronization. When a new command value is received, it cannot be responded to before the new duty cycle takes effect. This results in a time uncertainty that can last up to one PWM cycle (typically in the range of 50μs to 100μs). In fact, there will be an undefined and variable phase relationship between the network synchronization period and the PWM period. Compared to the time uncertainty of less than 1μs on real-time networks, it is clear that the motor controller's I/O plays a more crucial role in network-synchronized motion control. In reality, it is not the real-time network, but the system I/O that determines the synchronization accuracy.
Referring again to Figure 1, the system has three synchronization domains, A, B, and C, which are not bound together. They are not actually synchronized and have variable uncertainties that can last up to one PWM cycle.
Synchronization uncertainty and application impact
In high-performance multi-axis servo systems for applications such as robotics and machine tools, the impact of time uncertainty is clearly visible. Variations in the time offset between motor control axes at the I/O level have a direct and significant impact on the final 3D positioning accuracy of the robot or machine tool.
Consider a simple motion curve as shown in Figure 3. In this example, the commanded motor speed (blue curve) rises and then falls. If the ramp rate is within the capabilities of the electromechanical system, the actual speed will follow the commanded value. However, if there is a delay at any point in the system, the actual speed (red curve) will lag behind the commanded value, resulting in a position error Δθ.
Figure 3. Impact of timing delay on positioning accuracy
In multi-axis machines, the target position (x, y, z) is converted into an angular axis description (θ1, ..., θn) based on the machine's mechanical structure. The angular axis description defines a series of position/velocity commands with equal time intervals for each axis. Any timing differences between axes will lead to a decrease in the machine's accuracy. Consider the two-axis example shown in Figure 4. The machine's target path is described by a set of (x, y) coordinates. Delays cause timing errors in the y-axis commands, ultimately resulting in an irregular actual path.
In some cases, the impact of fixed delays can be minimized through appropriate compensation. However, more critically, it is impossible to compensate for variable and unknown delays. Furthermore, variable delays cause changes in control loop gain, making it difficult to adjust the loop to achieve optimal performance.
It should be noted that any delay anywhere in the system will lead to inaccurate machine precision. Therefore, minimizing or eliminating delays as much as possible is essential to improving productivity and final product quality.
Figure 4. Impact of timing delay on positioning accuracy
Synchronization and New Control Topologies
Traditional motion control methods are illustrated in the upper half of Figure 5. The motion controller (typically a PLC) sends position commands (θ*) to the motor controller via a real-time network. The motor controller consists of three cascaded feedback loops: an inner loop controlling torque/current (T/i), an intermediate loop controlling speed (ω), and another loop controlling position (θ). The torque loop has the highest bandwidth, and the position loop has the lowest bandwidth. Feedback from the factory remains local to the motor controller and is closely synchronized with the control algorithm and pulse width modulator.
In this system topology, the motion controller and motor controller achieve axis synchronization through position command values. However, in extremely high-precision applications such as CNC machining, the correlation with the motor controller's I/O (feedback and PWM) synchronization becomes problematic. Position loops typically have relatively low bandwidth and are therefore less sensitive to I/O synchronization. This means that even if the network and I/O are in different synchronization domains, the node synchronization performance at the command level is generally acceptable.
While the control topology shown in the upper half of Figure 5 is common, other control partitioning methods can also be used, such as implementing position and/or speed loops on the motion controller side and transmitting speed/torque command values over a network. Recently, the industrial sector has been moving towards a new partitioning method where all control loops are moved from the motor controller to a powerful motion controller on the network host side (see the lower half of Figure 5). The data exchanged on the real-time network consists of voltage commands (v*) from the motor controller and factory feedback (i, ω, θ) from the motion controller. This control topology, implemented with a powerful multi-core PLC and a real-time network, offers several advantages. First, the architecture is highly scalable. Axes can be easily added/removed without concern for the motor controller's processing power. Second, accuracy is improved because trajectory planning and motion control are performed at the same central location.
The new control topology also has drawbacks. The removal of the control algorithm from the motor controller results in a loss of tight synchronization between code execution and I/O. The higher the bandwidth of the control loop, the greater the loss of I/O synchronization. The torque/current loop is particularly sensitive to synchronization.
Figure 5. Traditional (top) and emerging (bottom) motion control topologies
Figure 6. The I/O scheduler binds synchronization domains together.
Recommended solution
Moving the faster control loop to the motion controller requires full synchronization from the network host to the motor terminal.
The overall idea is to synchronize all axis I/O with the network so that they all operate within a single synchronization domain. Figure 6 shows an I/O event scheduler located between the network controller and the motor controller. The main function of the I/O event scheduler is to generate synchronization/reset pulses for all peripherals, keeping them synchronized with network traffic. The I/O event scheduler acquires the frame synchronization signal, which originates from the network controller's local clock, and outputs appropriate hardware trigger signals for all I/O that must be synchronized with the network.
Each I/O has its own set of timing/reset requirements, meaning the I/O event scheduler must provide a customized trigger signal for each I/O. While the trigger signal requirements differ, they still adhere to some general rules. First, all trigger signals must be frame-synchronized. Second, there is a delay/offset associated with each trigger signal. This delay is related to the inherent delay of the I/O, such as the conversion time of an ADC or the group delay of a sinc filter. Third, there is an I/O response time, such as data transmission from the ADC. The I/O event scheduler understands the timing requirements of each I/O and continuously adjusts the trigger/reset pulses according to the local clock. An overview of the principle behind generating each output pulse of the I/O event scheduler is shown in Figure 7.
Figure 7. I/O scheduler generates trigger pulses
In most networked motion control systems, the frame rate and frame synchronization rate are equal to or lower than the PWM update rate of the motor controller. This means that the I/O event scheduler must provide at least one, and possibly multiple, trigger pulses per frame cycle. For example, if the frame rate is 10 kHz and the PWM rate is 10 kHz, the I/O event scheduler must provide one trigger pulse for each network frame; similarly, if the frame rate is 1 kHz and the PWM rate is 10 kHz, the I/O event scheduler must provide 10 trigger pulses for each network frame. This is equivalent to the frequency multiplier in Figure 7. A delay time tD is applied to each synchronization pulse to compensate for the inherent delay of each I/O. The final element of the I/O event scheduler is intelligent filtering. Some traffic jitter exists on each network. Filters reduce trigger pulse jitter and ensure that the rate of change of the frame synchronization frequency is limited.
The lower half of Figure 7 shows an example timing diagram for PWM synchronization. Note that in this example, the frame synchronization frequency is a multiple of the PWM frequency, and how I/O trigger signal jitter is reduced.
Implementation Plan
Figure 8 shows an example of a recommended synchronization scheme implemented and tested in a networked motion control system. The network host uses a Beckhoff CX2020 PLC and is connected to a PC for developing and deploying PLC programs. The real-time network protocol (red arrow) is EtherCAT.
The motor controller primarily uses Analog Devices' fido5200 and ADSP-CM408. Together, they provide a highly integrated chipset for network-connected motor drivers.
The fido5200 is a real-time Ethernet multiprotocol (REM) switch with two Ethernet ports. It provides a flexible interface between the host processor and the industrial Ethernet physical layer. The fido5200 includes a configurable Timer Control Unit (TCU) for advanced synchronization schemes against various industrial Ethernet protocols. Additional functions such as input capture and square wave signal output can also be implemented using dedicated timer pins. The timer inputs/outputs are in phase with the local synchronization time and therefore with network traffic. This allows it to synchronize not only the I/O of a single slave node but also slave nodes across the entire network.
The REM switch chip has two Ethernet ports, allowing connection of two Phys (PHY1 and PHY2). This topology supports both ring and linear networks. However, in this experimental setup, for demonstration purposes, only one slave node is used, and only one Ethernet port is active.
The REM switching chip communicates with the host processor via a parallel memory bus, ensuring high throughput and low latency.
The host processor used to implement the motor controller is the ADSP-CM408. It is a dedicated processor based on the ARM® Cortex®-M4F core, used for control and application functions. This processor includes peripherals supporting industrial control applications, such as timers for PWM inverter control, ADC sampling, and position encoder interfaces. To keep all peripherals synchronized with the network, a flexible trigger routing unit (TRU) is used. The TRU redirects the trigger signals generated by the fido5200's TCU to all timing-critical peripherals on the ADSP-CM408. These peripherals include pulse width modulators, sinc filters for phase current measurement, ADCs, and absolute encoder interfaces. The principle of synchronous I/O is shown in Figure 9.
Figure 9. Generating synchronization events for I/O
In Figure 9, note how the I/O event scheduler is implemented using the TCU on the REM switching chip and the TRU on the motor control processor. In other words, this function is implemented by two integrated circuits.
The motor controller provides feedback on the phase current and rotor position of the three-phase servo motor. The phase current is measured using an isolated Σ-Δ ADC, while the rotor position is measured using an EnDat absolute encoder. Both the Σ-Δ ADC and the encoder are directly connected to the ADSP-CM408, requiring no external FPGA or CPLD.
The PWM switching frequency is 10kHz, and the control algorithm is executed once per PWM cycle. As described in this article, the TCU provides a synchronization pulse to the ADSP-CM408 once per PWM cycle.
Experimental results
The experimental setup is shown in Figure 10. To illustrate the system's synchronization function, the PLC was configured to run a program task lasting 200 μs. The task duration also determined the frame rate on the EtherCAT network. The motor controller operated in PWM mode with a control update cycle of 100 μs (10 kHz), thus requiring synchronization pulses to be generated at this rate. The results are shown in Figure 11.
Figure 8. Implementation of the synchronization scheme
Figure 10. Implementation of the synchronization scheme
Figure 11. Generating synchronization events for I/O
The DataReady signal indicates when the REM switch chip should provide network data to the motor control application. The signal is set every 200μs, corresponding to the EtherCAT frame rate. A PWM synchronization signal is also generated by the REM switch chip to keep the motor controller's I/O synchronized with the network traffic. Since the PWM period is 100μs, the REM switch chip schedules two PWM synchronization pulses per EtherCAT frame. The two signals HSPWM and LSPWM at the bottom of Figure 11 are the high-side and low-side PWM signals for one of the motor phases. Note how the PWM signals are synchronized with the network traffic.
Summarize
Real-time Ethernet is widely used in motion control systems, and some protocols can achieve time synchronization with an accuracy of less than 1 μs. However, synchronization only involves data communication between the network master and slave devices. Existing network solutions do not include motion control I/O synchronization, which limits the control performance that can be achieved.
The synchronization scheme proposed in this paper enables end-to-end synchronization from the network host to the motor terminal. Due to the significantly improved synchronization performance, this scheme can substantially enhance control performance. The scheme also provides seamless synchronization across multiple axes. Axes can be easily added, and synchronization can be customized for individual motor controllers.
Synchronization is based on an I/O event scheduler located between the network controller and the motor controller. The I/O event scheduler is programmable in real time at high speed and can be adjusted to minimize jitter/frequency variation effects.
The proposed scheme has been validated in an experimental setup, and the results are demonstrated. The communication protocol used in the experiment was EtherCAT. However, the proposed scheme is applicable to any real-time Ethernet protocol.
References
1. Jie Ma, “Design and Implementation of a Multi-DOF Motion Control System Based on EtherCAT.” The 6th International Conference on Instrumentation, Measurement, Computer, Communication and Control, July 2016.
JensSorensen [[email protected]] is a Systems Applications Engineer at Analog Devices (ADI), responsible for motor control solutions for industrial applications. He holds a Bachelor of Science degree in Electrical Engineering from Aalborg University, Denmark. His primary interests lie in control algorithms, power electronics, and control processors.
Dara O'Sullivan [[email protected]] is a Senior Systems Applications Engineer in the Motor and Power Control (MPC) team of the Automation, Energy & Sensors business unit at Analog Devices (ADI). His expertise lies in power conversion and control for AC motor control applications. Dara holds a Bachelor of Engineering, Master of Engineering, and PhD degree from University College Cork, Ireland. Since 2001, Dara has worked in research, consulting, and industry on industrial and renewable energy applications.
Christian Aaen [[email protected]] is a Software Systems Design Engineer in the Deterministic Ethernet Technologies group at Analog Devices (ADI). His expertise lies in embedded software design, with a technical background in power conversion and motor drives. He holds a Bachelor of Science and a Master of Science degree from Aalborg University, Denmark.
Statement: This article is copyrighted by China Transmission Network. You are welcome to reprint it, but please be sure to indicate the source and author.
