fieldbus_tnail
fieldbus_tnail
fieldbus_tnail
fieldbus_tnail
fieldbus_tnail

Physical layer diagnostics: Who needs them?

Dec. 11, 2006
In this month’s installment of On the Bus, process control specialist John Rezabek explains the difficulties of getting a quality installation of physical-layer hardware for fieldbus diagnostics.
By John Rezabek, Contributing Editor

If you have any fieldbus segments in operation, there’s a chance you are still working on what to do with all the new diagnostics. There are diagnostics for every device, valves and positioners, most of the sensors and many system-level components. But until recently, there were none for the “physical layer”; that is, the power conditioners, cable, terminators and connectors that make the whole network possible.

One could always connect various testers or an oscilloscope and get a snapshot of segment health at that instant. While oscilloscope traces need some interpretation, testers like the Relcom FBT-3 or FBT-6 will do some of the diagnoses.

Aside from the “snapshot” nature of these measurements, there is also the work involved in connecting the test devices to 100 or 1000 or more segments on a routine basis. Most of us don’t have the luxury of performing such surveys with any regularity. Even when we do it “for cause” (e.g., mysterious communication errors in the night), connecting a given effect with any cause on the physical layer can be challenging. It’s even get more frustrating if we’re trying to convince a supplier of which problems are his or her responsibility and which are ours. Circumstances like these have driven many to seek a more “continuous” physical layer diagnostic.

Maris Graube, president of Relcom, is skeptical why anyone would need one. After all, says Graube, if you do the installation right, enforce a reasonable level of quality control, and apply a portable tester to each segment, you should be able to weed out all the installation issues that have the potential for future problems. Moreover, most systems have communication statistics on the host that further reveal possible segment errors.

But the difficulty of getting such a quality installation is reflected by Chuck Carter at Lee College, Baytown, Texas, who is designing a three-day course for installers. Maybe after crimping a few hundred ferrules, a certified installer can be let loose on your fieldbus (or conventional) plant installation. I think Chuck should have a room with a minus-20º wind chill to train his students for cold conditions, so we can use his initiates up here in the higher latitudes. Even in warm climates, Chuck has seen some real horror shows of loose terminations, shorted/grounded shields and missing or misplaced terminators.

In many places in North America, we are subject to whoever is signed up at the local union hall. Most of the electrical foremen here select their “terminators” with care. Not every journeyman passing through the IBEW hall is cut out to work gloves-off in the bitter cold, and achieve 99% or better accuracy.

Even if one achieves installation Valhalla, what about external conditions? Will an aging installation ever have a transmitter fill halfway with rainwater over the course of a season or two? Our friends on the Gulf Coast see their instruments and/or junction boxes submerged even when there isn’t a 20-foot storm surge.

Try as we might to attain a quality system, invariably some device is left with its plastic shipping plugs or unsealed metal ones in the unused conduit connection. What if a later project installs some variable frequency drives (VFDs) and the grounding of cable jackets proves to be insufficient? We have found birds, wasps and assorted other creatures who thought a warm instrument box was an ideal nesting place.

Even loose terminations may not show up until environmental conditions or a technician, reveals the errant crimp job. I remember a DP transmitter that only had errors when there was a stiff west wind. It sat on the west side of the unit, on grating and the right wind speed would get it vibrating on its pedestal. Only under those circumstances was it revealed that its factory-installed mini-fast connector’s leads had not been torqued to spec under the transmitter’s terminals.

Our supplier community has heard our worries, and we soon will have an interesting selection of new utilities for gaining insight into the health of the physical layer. It appears the cost of the advanced physical layer diagnostics will range from $300 US per segment and up, so users will have to contemplate whether the problems they experience (or fear) are widespread enough to warrant full coverage.

Those with compact plants, less subject to hostile environments, may choose a small number of such modules placed strategically and/or rotated, until they have enough history to judge whether more coverage is needed. At our plant, we anticipate a gradual move toward 100% coverage of all “level 1” and “level 2” criticality loops, depending on costs and accessibility (unit outages).

In next month’s column, I’ll summarize the current and soon-to-be-released offerings of four of our leading suppliers of physical-layer hardware.

  About the Author
John Rezabek is a process control specialist for ISP Corp., Lima, Ohio. Email him at [email protected].