By Mike Bacidore, Managing Editor
The significant upside of the diagnostics, maintenance and troubleshooting benefits that a digital network provides is all too alluring to ignore. Digital diagnostics provide more and better information that helps to troubleshoot and fix problems, and without shutting down the entire system.
Eric Rice, director of technical support for North America at FKI Logistex (www.fkilogistex.com), Cincinnati, which designs and builds conveyance and palletizing machines, agrees that, from a cost standpoint, the digital network makes a lot of sense.
"You don't really gain anything in speed because the digital network is a little slower," he explains. "But the amount of diagnostics you have is better. If you have an issue in one part of the system, it doesn't mean the entire system is down. Subsystems can run independently, and you have more uptime. In today's conveyor system, digital networks are the lifeblood."
With the unstoppable wave of digital adoption, diagnostics have changed, but only slightly. "Our products' digital signals consist of 24 Vdc I/O signals and Ethernet I/P," says Wade Peterson, electrical engineer at CMD, builder of bag- and pouchmaking equipment in Appleton, Wis. "Typically, digital signal problems for us are rare as our machines have a very small footprint and the cables are well-shielded and isolated from noise. Most problems are typically wiring-related like a broken conductor or nicked insulation. Our troubleshooting methods therefore focus on the media first."
The best first step could be going a visual inspection, says Jack Chopper, chief electrical engineer at Filamatic (www.filamatic.com), a liquid filling and packaging machine builder in Baltimore. "Fortunately, we don't encounter very many digital signal problems, but we'll use network analyzers and testers, both hardware and software, built-in diagnostic tools, recording instruments, multimeters and scope meters."
The tools of choice seem consistent from machine builder to machine builder. "We typically troubleshoot with a combination of multimeter, oscilloscope and PC-based tools," adds Peterson.
From a discrete troubleshooting standpoint, the most frequently used equipment at FKI Logistex is a multimeter. "Those kinds of checks are done on-site," says FKI Logistex's Rice. "From a network point, different buses have different values we look for. For a Profibus network, we have a Profibus monitor. We look at the integrity of the network. We can do the same thing for Ethernet. We also have some proprietary networks that have resistance values. If it's out of specification, you may see intermittent issues with that piece of equipment. We have Ethernet sniffers—software or hardware—we would use to troubleshoot or get statistics."
At MAG Americas (www.mag-ias.com), it's done in two different ways. "In the machine realm, for qualification, some tools are provided by our manufacturers like Fanuc or Siemens," says Jim Braun, vice president, product development and standardization for MAG, a large machine tool and systems company in Hebron, Ky. "We have other devices for tuning. We have Heidenhain scales. We have devices we can insert in the signal path or in the network—Profibus or Ethernet. That's primarily on the machine qualification side. In the field, we have internally based diagnostic tools that run on the controls and give some pass/fail indications. We're also looking at some other more advanced tools we haven't released yet, like having a built-in scope feature inside the control." MAG uses a lot of Profibus and Ethernet-based interfaces that are proprietary to the control, and other parts of the machine may take special types of equipment, too (Figure 1).
"During the building and qualification of the machine, we have some tools to debug the machine," explains Braun. "Once the machine is finished and it ships, if there would be a problem, we have remote diagnostics, but it's only available on PC-based controls where we can request control of the machine and the person on-site can give us control."
Most digital signal protocols include some form of error-checking or diagnostic information, explains Mara White, industrial Ethernet marketing manager at Fluke Networks (www.flukenetworks.com). "The more intelligence or sophistication built into the diagnostics, the more complex the measurements and analysis algorithms become," says White. "This has an exponential impact on cost and can even lead to compromises in network performance. It is the compliance to digital protocol standards and the built-in diagnostics and error-checking that allow machine builders to integrate various subassemblies. This can be very effective for the non-custom jobs but also protects them when the line of transfer of ownership becomes an issue."
When integrating multiple machines from a variety of vendors, use a combination of built-in machine diagnostics, diagnostic tools and appropriate maintenance strategies to ensure optimum system performance, says White. "Relying solely on the machine's diagnostics for equipment breaks can be a risky proposition," she warns (Figure 2). "Testing and troubleshooting comprise a small cost that can save numerous hours and headaches during the initial troubleshooting and ongoing maintenance of the machine and network. And the value increases with the mission-critical and time-sensitive nature of the work."
Did You Check the Cable?
The best way of dealing with any troubleshooting issue is to not have the problem in the first place. "This may sound strange, but many problems are ‘designed-in' by users not following the specifications of the digital network," says Helge Hornis, PhD, manager, Intelligent Systems Group, Pepperl+Fuchs (www.pepperl-fuchs.com). "For instance, a network where a segment must not be longer than 100 m—examples are Ethernet and AS-Interface—shouldn't be set up to have 200 m cable. But if trouble somehow shows up, a combination of an oscilloscope that allows evolution of the waveform on the network and a dedicated analyzer that allows evaluation of the data packets traveling over the network is best."
Fluke Networks' White recommends dissecting the node or segment into its finite elements, starting from the end device on the segment and working backward to the controller. Often an up-to-date blueprint can simplify this task, but the main elements can include the physical medium, signal properties such as amplitude, timing or baud rate and communications protocol.
"Inspect and check every connection, termination and cable for signs of wear or damage," suggests White. "Pay attention to cable installation length, connector wear and corrosion and even quality of cables and connectors. Use an appropriate measurement tool to validate the cable properties are within the manufacturer's or installation specifications. Critical parameters include impedance, capacitance, length, connector wiremap, cross talk and noise. It's important to know if the cable installation can support the intended network communication performance rating. There are a variety of cable analyzers on the market measuring anything from basic continuity or application bandwidth to fully comprehensive detailed diagnostics."
Once the cable infrastructure is known to be good and within specification, the next step is to verify the electrical integrity of the digital signaling. Analyzing the signaling integrity can be broken down into two areas. "Check amplitude, frequency or transmission rate characteristics," says White. "Taking measurements at critical points across the entire segment can reveal signal attenuation or sensitivity problems or device setup conflicts. Then check for signal distortion, caused by cable impairments such as transmission signal reflections or external influences like induced EMI, such as line power interference, random high-voltage transients or static discharges." An oscilloscope with appropriate measurements and waveform capture and analysis capabilities can provide a visual insight into the quality of digital signals, she suggests.
Analyzing the network communications has often been seen as the role of the IT specialist, but, with TCP/IP-based networks rapidly being deployed in the industrial world, the maintenance professional is being asked to learn new technologies, techniques and tools. "Troubleshooting can be as simple as looking at the obvious, like network activity icons or indicators," says White. "Non-invasively probing into the network and monitoring critical performance factors can lead you to localizing the problem. And testing an installation to validate whether or not it can support higher bandwidth or faster traffic can determine if production process speed improvements can be realized. A portable network protocol assistant, with an Ethernet, wireless or fiberoptic interface, can provide the measurements to diagnose TCP/IP-based problems on industrial networks."
One more practice White recommends is baselining—documenting critical performance measurements at installation, before and after repair or when changes are performed. "Keeping accurate baseline measurements will allow you to compare and contrast tests, giving you the ability to quickly identify differences, thereby rapidly isolating problems," she says. "Network cable infrastructure, electrical signaling and communications protocol performance issues can, over time, translate into real production losses in the form of downtime or inefficiencies."