Same 50 faults, just never in the same order? Try CIP Sync
|Traditional alarming systems record and log the time of failures in serial order based on when the HMI receives the data. That means alarms triggered 100 milliseconds after the root cause could get time stamped as occurring before the root cause, simply because the data was received first.|
It’s 3 a.m., and, for the second time today, line four has experienced a major shutdown in your manufacturing production line. The results are paper jams in subassemblies one and two, a torn web in two places, and 100 parts in process that need to be scrapped. Even worse, machines are halted, motion controls aren’t moving because glue gun three is dumping adhesive onto the belt because the machine fault is on top of an on-position for the gun — in the precise spot that the web broke.
You’re in bed when the third-shift production manager calls. The restart procedure and glue cleanup take about an hour, while upstream lines that feed material into line four have accumulated inventory stacked on the floor waiting to be processed. Downstream production, of course,
Tos to show the same faults in a different order each time. It always reports the same 50 faults, just never in the same order.
Limits of traditional
All of the faults occur in such rapid succession, you just can’t resolve which came first. You’ve tried to adjust all of the clocks in your architecture manually. You‘ve even implemented a Network Time Protocol (NTP) interface to the controllers in an attempt to increase the time stamps’ accuracy. But, these events occurred only 10 to 20 milliseconds apart, and the NTP system cannot synchronize the controller clocks that closely. Moreover, the poll cycleon your alarm and event system interjects 250 milliseconds of error by itself.
Figuring out and addressing the root cause of the fault becomes critical in situations like this one. Unfortunately, the typical diagnostic system available today cannot easily identify the root cause of failures.
Traditional alarming systems record and log the time of failures in a serial order based on when the HMI receives the data. That means alarms triggered 100 milliseconds after the root cause could get time stamped as occurring before the root cause, simply because the data was received first.
CIP Sync, a time synchronization extension to the Common Industrial Protocol (CIP) managed by ODVA, can help solve this problem. CIP Sync forms the heart of the EtherNet/IP Time Synchronized Distributed Control approach for motion applications. Additionally, CIP Sync is based on and compliant with the IEEE 1588 Version 2 standard for a precision clock synchronization protocol for networked measurement and control systems. It allows for high-precision time synchronization so that I/O or controllers can provide very accurate first-fault detection.
CIP Sync provides a standard mechanism to set and synchronize the clocks on multiple controllers throughout the control architecture. Some components can be coordinated down to an accuracy of 100 nanoseconds. Additionally, optional global positioning system (GPS) interfaces help to make sure that a high level of accuracy can be obtained across geographically separated areas of automation or substations.
Also, CIP Sync uses the Precision Time Protocol (PTP) to distribute Coordinated Universal Time (UTC) across a standard Ethernet network. UTC is an offset independent time in the controller (no time zone or daylight savings time offsets). By time stamping in UTC, events can be compared across time zones without the need to adjust for the geography in which they were generated. In addition, UTC is immune to local daylight savings time, which can corrupt a database or sequence of event (SOE) application. Time zone and daylight savings time offsets are then added back into the time stamp automatically when displayed to an operator by a client application, such as an alarming system.
|CIP Sync allows you to schedule outputs, useful in diverter applications to trigger multiple outputs simultaneously or to trigger a reject at the precise moment at which a product is at a reject station. Source: Rockwell Automation|
Within an integrated control environment, such as the Rockwell Automation Integrated
Simplified controller clock settings. Large plants often have more than 1,000 controllers, all with wall clocks that need to be maintained. CIP Sync uses coordinated system time (CST) to streamline wall clock setting to offer high-precision time stamping and coordinate the wall clocks. Simply set one master clock and all the controllers will report alarms and events using the same system time.
Database integrity. Many industrial applications such as track and trace, first-fault detection, and event tracking require that large volumes of historical data be stored in a central database. All of this data is meaningless, however, without a common reference as to when it was generated. CIP Sync creates a common logging time between systems.
Position registration. Most networked I/O cards cannot send registration on/off information quickly enough for the controller to accurately generate an axis position. With CIP Sync, time stamps made directly on the I/O card establish when the mark was in front of the photo eye. A lookup function in the motion planner then returns axis positions that equate to the registration time, allowing a single time stamp to be passed to all axes in the system and returning position.
Improved fault tolerance. CIP Sync is a multimaster time network that self-arbitrates to identify the best time source in the control system, providing a more fault-tolerant architecture.
Dave Rapini is
For more on CIP Sync, see:
> ODVA announces new editions of CIP network specifications and testing of Ethernet/IP products
-Edited by Renee Robbins, senior editor
Control Engineering News Desk, www.controleng.com
Register here to select your choice of free eNewsletters.