Advantages of Remote I/O Devices With Networking Technologies

If the average machine designer were given the option of setting up the specs for the ultimate 'dream machine' operation, it would most certainly include every modern convenience, every networking option, and every high-efficiency feature available on the market. That may sound great today, but in a couple of decades, that modern technology will become obsolete.

How can you plan for the future when you don’t even know what will exist?

There is no way that you can predict twenty, forty, or fifty years into the future to make life easy for the next engineer by planning a system upgrade. But you can make it easier for them now. How do you do this? Design it to be modular.

The natural inclination of electronic technology development is to examine the demands of a machine and assign each input/output device its own set of wires. Those wires trace back to a controller for the system as dictated by a ladder schematic illustrating the entire machine center in impressive complexity. But this myriad of wires can come with a few major disadvantages.

Figure 1. A typical remote networked input/output system. Image used courtesy of Adobe Stock

Disadvantages of Direct I/O Wiring

There are three primary problems that can affect the direct wiring of discrete I/O devices to a controller located more than a few hundred feet away (sometimes even less).

First, the resistance value of long, straight wire runs creates a voltage drop that can render a signal unreadable. If you need 24 volts at the controller, the wire itself may drop too many volts, and the input voltage signal will fall below the minimum threshold. Therefore, you do not have the option of delivering that signal a couple of hundred feet away.

A second disadvantage is that of capacitance. Two conductors running in the same duct, conduit, or cable tray will impose a capacitance on each other, and this can negatively affect I/O signals that rapidly switch on and off. For slower signals, it may not be as much of a problem (remember the current across conductors for capacitance is related to 2*pi*f*C, so a higher switching frequency imposes more current between wires).

The final disadvantage is not a technical characteristic of voltage, but rather the complexity of wiring. When considering long bundles of wires that span between machines, there are many headaches and frustrations that come along with even a single point of failure in the wiring. Although simple I/O devices may be less expensive, the time it takes to integrate these devices with the rest of the system, and then to locate and solve failures later, can be a heavy cost burden.

So, for these reasons, we might opt for either a separate PLC system onboard each machine or simply resign ourselves to the complex control wiring with the rest of the system. Neither one is a favorable option for the technicians who will be servicing the machine in future years.

Remote I/O Provides Flexible and Simple Solutions

Nothing makes an engineer or technician happier than a machine that costs less to install, works better than the old machine, and is easier to operate and repair than the previous installation. That seems too good to be true, but it's one of the main purposes of segmented computer networking.

Networking cables are capable of sending complex streams of data over 300 ft in the case of standard network cables, and much longer in the case of fiber optic cables. Complex combinations of digital and analog input/output data can be sent over hundreds of feet very quickly. The concept of remote I/O is simply to place a distribution point at the end of these network cables which can collect sensor signals, then transmit them over the network to a single processor. They can also receive commands from that center processor and complete tasks of their own.

Figure 2. The typical structure of a simple remote I/O network with a PLC. 

This capability allows us to build a giant network of sensors and output devices, and operate them over long distance network cables, all being operated from a single central processor. 

An important consideration would be to establish a very consistent addressing system that assigned each group of machines with its own set of network addresses (IP Addresses for an Ethernet system). This allows the technicians to track down problems very quickly based on device address, and also allows rapid deployment of new machine centers simply by adding another network distribution point and set of network addresses. Establish that new single network connection with a cable, and the entire machine center is operational - rapid deployment!

I/O System Hardware

The first requirement is a process controller or programmable logic controller with a network connection. A remote I/O network can only send and receive information, it can’t make any decisions. Think of this as trying to use a keyboard without attaching it to the computer. The keys are pressed, but without the computer to process anything, the button pressing won’t do anything useful.

In a similar manner, I/O networks will only operate with the main controller to handle the logic programming.

Usually, this will be a PLC with an Ethernet or similar connection. Each of the networked or remote I/O chassis contains the network bus and a number of modules with wiring points for the sensors, relays, and other I/O devices.

Some of these systems will only need a half dozen I/O points, but others may have well over 100, it all depends on the scale of the machine. Usually, these points will take the form of screw terminals for +24, GND, and the signal. Some modules provide access to 5-pin M12 quick-disconnect cable terminal.

Figure 3. A Murrelektronik remote I/O module, rated for machine mounting conditions, is on shown display at a trade show.

Adding the I/O Devices to the Project

The exact method for adding the devices to the project will be up to the hardware manufacturer and the software for the PLC. Usually, the only required information is the IP address of the remote I/O network module as well as the information about each I/O module.

For most networked remote I/O modules, the manufacturer provides an electronic data sheet (EDS), general station description (GSD), or EtherCAT subdevice info (ESI) files that can work with networks across multiple manufacturers, a massive benefit for retrofitting systems. 

With this information, you add the device into the project, then each remote I/O terminal is accessible with its own address and tag, just as if it were connected directly to the PLC itself.

Many times, it can be possible to add an entirely new machine center into the operation with the only software change being a few lines to activate a new sub-routine of programming. Apply power to the machine center, connect a network cable, and the new devices are operating as intended.

Even if this new hardware doesn’t include an entirely new machine center, the concept of remote I/O can still make a major impact in your operation.

Imagine adding a PLC with a network connection to the engineer’s office. Out on the floor, add a network I/O hub to each machine with a single sensor giving the status of each machine. With just a glance, the engineer can see the production of each machine, the trends of the past day and month, and can calculate the exact costs of downtime and maintenance for each machine over a given time span.

With this new information, the engineer can give priority to the maintenance of each machine and can predict when to take corrective action to prevent failures.

The future of automation challenges us to design with flexibility and expandability in mind. With powerful computers running a process, we no longer need to devote a separate control system for each machine. Networked remote I/O allows us to design a machine center separately from the rest of the system, and drop the solution right in place.

It can also be used to upgrade and add I/O into existing systems for far less cost and difficulty than ever before.

Original article published Feb 2020