- What is POWERLINK?
- Why should I use POWERLINK?
- What are some applications where it makes sense to use POWERLINK?
- What types of devices can take advantage of POWERLINK?
- What kind of performance can I achieve with POWERLINK?
- Is POWERLINK an international standard?
- How does B&R ensure that all POWERLINK devices conform to the standard?
- POWERLINK is a complete open technology. What does this mean?
- What are the advantages of using POWERLINK?
- POWERLINK is known as CANopen over Ethernet. What does this mean?
- Why is POWERLINK ideal for an "integrated automation" concept?
- How can I troubleshoot POWERLINK network?
- How can I calculate the POWERLINK cycle time?
- I have done a network trace that shows non-POWERLINK frames. Is something wrong?
- My network trace shows some frames coming in the wrong sequence, but no error is detected by the master. Is something wrong?
- How do I find sporadic errors in a POWERLINK network?
- I have some problems with my slave devices. Whom can I contact?
- I have a question about the technology itself. Whom can I contact?
- What is the available asynchronous bandwidth?
- What is the difference between static and dynamic mapping?
- My network programming/configuration tool does not accept my device description file (xdd). What can I do?
Ethernet was originally designed for office applications and is not deterministic (due to CSMA/CD, latency, queuing in switches, etc.). To satisfy the requirements of industrial automation and process control, the concept of real-time industrial Ethernet was introduced to the market. POWERLINK, a software protocol based on standard Ethernet, is the most well-known of these real-time industrial Ethernet solutions. High-speed, deterministic response times are ensured via through a mix of time slot and polling procedures.
If your application requires high-speed, deterministic response times, POWERLINK is the right solution for you. POWERLINK is based on standard Ethernet and does not require any specific proprietary hardware (ASICs, etc.). POWERLINK boosts your application to unmatched performance levels at the lowest cost.
POWERLINK is used in all kinds of applications in all industries, including audio/video, automotive, distributed control systems, energy management, machinery, industrial automation, railway and maritime transportation, robotics, uninterruptible power supplies, vision systems and many more.
There's only one reason not to put a POWERLINK interface on a device: "If the device is too small for an Ethernet port". For all other devices, it makes sense.
Cycle times down to 100 µs can be achieved with 100 ns network jitter. The minimum cycle time depends on the amount of data, the number of nodes in the network and the frequency at which they need to communicate.
Yes, POWERLINK is defined as industrial Ethernet standards IEC 61158-13 and IEC 61784-2. POWERLINK is the highest graded industrial Ethernet standard in China (GB/T 27960-2011) and therefore recommended by official authorities.
To maintain POWERLINK as a global standard and ensure complete interoperability, B&R conducts certification events on a regular basis.
It means that POWERLINK technology is a standard technology widely used around the world that is independent of any country or company. Anybody can use and implement POWERLINK without a license or patent associated with the technology. In addition, an open-source POWERLINK software stack is available at no charge at https://sourceforge.net/projects/openpowerlink/
POWERLINK brings high performance and real-time capabilities to your industrial communications while reducing the design and operational costs of your application. The use of standardized device profiles ensures complete interoperability and your choice of POWERLINK devices. For users and OEMs, additional benefits of POWERLINK include maximum performance, increased productivity and easy integration as well as the advantages associated with reduced downtime, a fully open technology and troubleshooting features. For component manufacturers, no specific hardware is required; it's also a key opportunity to enter the largest established market. Training and worldwide support are also available.
POWERLINK integrates the entire range of CANopen mechanisms (object dictionary, PDOs, SDOs, etc.) The devices profiles used by POWERLINK have been defined in cooperation with CiA (CAN in Automation) and are therefore the same as CANopen device profiles.
The concept of "integrated automation" refers to modern machinery designs in which all devices are connected to a network that acts as a backbone for the machine. Since each device type comes with its own constraint on the backbone, only a flexible and decentralized technology can fulfill the necessary requirements. For example, motion and robotics applications usually require very low jitter and short cycle times, vision systems result in high usage of non real-time bandwidth, sensors will use a few bytes of data and heterogeneous cycle times, etc.
Thanks to the following key features, POWERLINK is ideally suited for integrated automation:
- Direct slave-to-slave communication (fastest drive-to-drive response time, no need to route through the master)
- Multiplexed slot assignment (no need to exchange all data at the fastest cycle time)
- Dedicated asynchronous phase within each cycle (large non real-time bandwidth)
- Poll response chaining (large amount of devices with few amount of bytes)
Because it is a software-based protocol, POWERLINK can be integrated into any hardware platform available on the market.
Nevertheless, there are some things to pay attention to when deciding on the platform. With a master implementation, for example, the operating system plays a crucial role. Cycle times down to 250 µs with very low jitter are easily achievable using a Linux operating system running on a standard PC. For slave implementations, in contrast, hardware acceleration is recommended in order to assure the response time of 960 ns of the slave device in a POWERLINK network. Most platforms already offer full support for this type of hardware acceleration.
There are three ways of receiving the stack:
- Technology integrators Ixxat, Kunbus, Port and Softing offer the stack for several hardware platforms in addition to services directed towards integration.
- Hilscher and HMS have a different approach to the market. They sell the hardware platform that contains the protocol stack as a black box and offer various integration services.
- Since POWERLINK is open-source software, its source code is available on the SourceForge code repository at https://sourceforge.net/projects/openpowerlink/. In this case, support is provided by an online forum.
Normal 0 21 false false false DE ZH-CN X-NONE </xml> POWERLINK can run on any operating system. The cycle time and jitter of the network are directly related to the operating system itself. A Windows-based POWERLINK master can achieve cycle times around 10 ms, whereas a Linux-based POWERLINK master can run at 250 µs.
Because POWERLINK uses single frames to transmit data over the network, the network cycle can easily be adapted to the needs of the application. All features are compatible to each other and can be used in networks that also contain devices that may not support those features:
Multiplexing: This feature makes it easy to combine high-frequency motion applications and temperature control in a single network. Axes that exchange positioning information in each network cycle are combined with temperature sensors that are multiplexed every nth cycle.
Poll response chaining: When you design your system with POWERLINK you don't have to decide whether to go with a centralized or distributed design. Poll response chaining makes it possible to respond to slaves in a network based on timing.
MultiASnd: Cameras are widely used in industries such as mobile machinery. Nevertheless, cameras produce a lot of asynchronous traffic on a network. In the past, this required extra wiring to separate real-time and non real-time data. With POWERLINK, you can combine both types of data in a single network. MultiASnd increases the asynchronous bandwidth to the extent necessary for this.
Redundancy: Redundancy is crucial in process automation applications as well as in the wind industry. Failure in a plant or a wind turbine results in soaring costs, which highlights the importance of reducing downtime across the board. With POWERLINK, you can choose the type of redundancy that keeps your downtime closest to zero, whether it's ring redundancy, cable redundancy or master redundancy.
Setting up a POWERLINK master configuration is simple. Each slave component has an associated XML device description file (XDD). Several engineering tools are available to configure the network (e.g. openCONFIGURATOR in the SourceForge code repository). These tools allow you to set up the network and create the relevant variables visually. Third-party products can be imported and integrated using XDD files.
After setting up and configuring the network, the tool automatically generates a CDC file that serves as the POWERLINK master configuration file.
Each POWERLINK slave comes with an XML device description file containing all of the parameters and variable information required to configure the slave. Once imported, setup and configuration are done as usual.
PLCs running as POWERLINK masters often have a logging mechanism used to report errors. Check the log file for any entries that refer to POWERLINK.
POWERLINK itself also keeps error counters on each device. The counters count loss of frame errors, CRC errors, timeouts, etc. and can be read from the corresponding objects via SDOs.
In addition, any standard network tracing tool can be used to perform diagnostics at the network level. The two most commonly used tools are Wireshark (open source) and OmniPeak (commercial). You just need to connect your PC to the network and run the tool in order to diagnose the traffic. You should connect the PC to the POWERLINK network via a hub and disable all protocols (TCP/IP, etc.) on the tracing interface. For high-precision network traces, the use of Ethernet analysis hardware is recommended.
The minimum possible POWERLINK cycle time is determined by the number of nodes on the network and the amount of data transmitted to and from each node.
Each network has the following frames:
- 1x SoC (64 bytes)
- nx PReq/PRes (64-1500 bytes)
- 1x SoA (64 bytes)
- 1x ASnd (300-1500 bytes)
Transmission time is calculated by the total amount of bits on the network (payload bytes * 8 + 64-bit header), determined by (bits * 10 ns). The response time also has to be added (ideally 960 ns gap between all frames).
On a standard PC, network interfaces are usually configured for TCP/IP networks to establish communication channels to other PCs or the Internet (e.g. DHCP for IP address requests, SMB traffic to find Windows shares, HTTP, etc.). This traffic is visible in the network trace and may also influence your POWERLINK network.
As a result, all protocols should be disabled on the interface you are using for the network tracing. How to do this depends on the operating system. In Windows, you can disable all protocols in the network card's adapter settings.
It is very likely that there are problems with the network trace. In order to get a good trace, you should use a separate PC connected via a hub to the other devices and disable all protocols on the tracing interface. For deterministic tracing with precise time stamping, the use of a network analysis tool is recommended.
As a first step, you should isolate the problem and make it reproducible. Use of filtering function of network traces will help localize and identify the error. With some network tools it is possible to enable triggering functions, which help trace collection when the sporadic error appears.
Your first contact will be the support team of your device manufacturer. In order to obtain help efficiently, prepare information such as the device type of the slave, the corresponding XDD file, the type of POWERLINK master (PLC type, operating/ runtime system version), any error logs from the device or master and, if possible, a Wireshark trace indicating the problem.
The amount of data transmitted in one asynchronous frame is determined by the asynchronous MTU (maximum transfer unit). This value can be set in the network configuration tool on the master side. Changing the value can allocate additional asynchronous bandwidth to achieve higher throughput. In addition, you can use the "Multiple ASnd" feature, which allows more stations to send asynchronous traffic in a single cycle.
PDO mapping determines which data is cyclically transmitted to and from the network. If a device uses static mapping, the mapping configuration is specified by the device manufacturer and cannot be changed by the user. If a device uses dynamic mapping, the configuration can be set by the user according to the application's requirements.
Typically, devices with little cyclic data, such as encoders or sensors, use static mapping, while devices with lots of settings such as drives use dynamic mapping.
You can validate your XDD file using the online XDD-Check utility on this website. If the checker reports errors, get in touch with the device manufacturer who provided the XDD file. If the checker cannot find any errors, get in touch with the provider of your configuration or programming tool.