Introduction to Dual Ethernet Architectures

National Instruments introduces a new family of Programmable Automation Controllers (PACs) that offer dual Ethernet ports across the CompactRIO, Compact FieldPoint, and Smart Camera platforms. With the power of dual Ethernet ports, these real-time controllers can separate your industrial I/O communication from your public enterprise network, improving data integrity and bandwidth usage.

Introduction

Ethernet provides the backbone for business, home, and person networks, and its broad adoption results in a cost-effective and reliable networking solution in industrial applications.  With the power of Ethernet, National Instruments Programmable Automation Controllers (PACs) bring networked I/O to industrial monitoring and control systems.  These controllers allow for programmatic communication over the network, automatically publish measurements over the network.  With features like LabVIEW Remote Panels, Web servers and built-in FTP/HTTP servers, you can access I/O data nearby in the same office or miles away from a remote, harsh environment.

Adding Dual Ethernet Ports

Ethernet is a popular networking choice because of the availability in the business and enterprise.  However, Ethernet offer the data integrity and security needed in industrial settings?  How does one dependably collect measurements and control complex systems without fighting for shared network bandwidth or adding extensive overhead?  To enhance the reliability of the Ethernet I/O platform, National Instruments has recently introduced PACs with dual Ethernet ports and high end PowerPC processors (see Table 1).  Now you can use these rugged real-time controllers on two separate networks: one port for network communication to a host PC or enterprise system and the other port for expansion I/O to other Ethernet-based devices.

Table 1. NI Real-Time Controllers with Dual Ethernet Ports

NI cRIO-9074 Integrated 400 MHz Real-Time cRIO Controller and 2M Gate FPGA	NI cFP-2220 400 MHz Real-Time Controller for Compact FieldPoint
NI 1742 Smart Camera 533 MHz Monochrome VGA Smart Camera with Built-in Lighting Controller	NI 1722 Smart Camera 400 MHz Monochrome VGA Smart Camera

Why Dual Ethernet Architectures?

The primary advantage of dual Ethernet ports is the ability for a real-time controller to create its own secondary private network, separate from any public network exposed to the Internet.  By isolating devices from the public enterprise network, the entire Ethernet bandwidth on the private network is dedicated to the I/O and HMI data you wish to pass through, improving response time and decreasing latency caused by busy traffic.

NI Dual Ethernet Technology

Each Ethernet port on the NI real-time controller has all the usual functionality of a network card.  For example, both ports communicate over Ethernet protocols such as TCP/IP, UDP, and Modbus TCP.  The dual Ethernet ports on CompactRIO and Compact FieldPoint offer 10/100 Mb/s transfer rates, and the Smart Camera provides two Gigabit Ethernet ports (10/100/1000 Mb/s). 

By default, Shared Variables and the FieldPoint API are hosted out of the primary port (Port 1). Shared Variables can also be hosted out of the secondary port (Port 2), but not both at the same time.  For example, Shared Variables hosted on the cRIO-9074 controller can be served out through Port 1 or Port 2 to devices connected on those separate subnets, but those Shared Variables cannot be served to both subnets at the same time.  The setup in Figure 1 shows a host PC passing its Shared Variable 1 to the cRIO Port 1, and the cRIO controller serves up its own Shared Variable 2 through cRIO Port 2 to the Compact FieldPoint system.

Figure 1. cRIO-9074 System Hosting Shared Variables.

Because the Shared Variable cannot be hosted out of both ports of the controller, connecting multiple dual-port devices in a long daisy-chain can cause unnecessary complication in programming.  A switch is recommended to pass data from one controller to several other controllers.

For all of these controllers, the software configuration utility Measurement & Automation Explorer is used to set the IP address for both ports.  For the NI Smart Camera, the NI Vision Builder for Automated Inspection software can also be used for configuration.  Step-by-step instructions and tutorials are provided below:

Related Links:
How do I Configure the Dual Ethernet Ports on Real-Time Controllers?
Getting Started with Dual Ethernet Port Real-Time Controllers
Deploying Shared Variables to a Specific Network Card
Using Shared Variables on the NI Smart Camera in Vision Builder AI

Example Architectures

When designing the architecture for networked systems using dual Ethernet ports, consider the example architectures provided below.  Figure 2 shows a typical set-up using 1 controller with dual Ethernet ports.  Port 1 (or Subnet A) on the controller can support DHCP or static IP addresses, whereas Port 2 (or Subnet B) can only be set to a static IP address.  Also, the Private Network Target in Figure 2 refers to any I/O or HMI device with at least one Ethernet port.  If the target is another real-time controller, remember that only one port on each controller can host Shared Variables and the FieldPoint API.

Figure 2. Example Architecture for a Single Dual Ethernet Port Controller.

For applications involving more than one controller or device with dual Ethernet ports, Figure 3 illustrates a viable architecture that uses switches for connecting all Port 1’s to Subnet A and connecting all Port 2’s to Subnet B.  All the attached Ethernet devices could be remotely viewed and controlled through the Internet or enterprise network on Subnet A, whereas Subnet B would be used separately among the devices as their dedicated network for passing I/O data back and forth.  Plus, connecting every device to both subnets brings the added advantage of easy IP address configuration management by the host computer.

Figure 3. Example Architecture for Multiple Dual Ethernet Controllers and Devices.

Use Case 1:
An example use case for the dual-port cFP-2220 controller would involve connecting a host PC to Port 1 and the cFP-1808 expansion chassis to Port 2.  Since the cFP-1808 is not a real-time controller, it uses the FieldPoint API, instead of Shared Variables, to serve data out its single Ethernet port.  Therefore, the cFP-2220 controller can freely pass data between the host PC and the cFP-1808 by using its primary port to host Shared Variables.

Use Case 2:
Another example involves using the cRIO-9074 as the dual-port controller.  Port 2 uses a switch to link to various PACs and I/O devices that communicate using Shared Variables, FieldPoint API, and Modbus TCP.  The cRIO-9074 serves as a central database for processing and displaying the data on an Ethernet-based HMI connected to Port 1.

Conclusion

Building upon the advantages of Ethernet networks, the National Instruments PAC platform now offers dual Ethernet ports for industrial monitoring and control systems.  By simply isolating the measurement and HMI data with these rugged real-time controllers, you can easily control bandwidth usage and maintain data integrity on this dedicated secondary network.  Separate your enterprise and expansion I/O networks to bring increased reliability to your industrial Ethernet-based applications.

Related Links:
Getting Started with Dual Ethernet Port Real-Time Controllers
NI cRIO-9074 Product Page
NI cFP-2220 Product Page
NI Smart Camera Product Page

6 ratings | 4.17 out of 5

Read in English Chinese (Traditional) Japanese Korean | Print

Reader Comments | Submit a comment »