CompactLogix Behavior Case Study

THE MIGRATION STORY

On a project to migrate a control system from multiple 1746 (SLC 5/03) PLCs and I/O to a centralized CompactLogix PLC with a distributed I/O system we chose to implement the Rockwell Automation SLC 500 to 5069 CompactLogix Conversion System. This system has the potential for saving significant time in system migration. This is particularly the case when the existing system has minimal or non-existent documentation.

Rockwell’s Integrated Architecture Builder (IAB) tool makes migration straightforward. In many cases the system designer will find direct replacements for the 1746 series I/O in the 5069 family. The simplicity can hide design parameters that should be considered when deploying the hardware.

Having a drop-in replacement system like this is a great time saver. However, there is at least one additional design parameter the engineer should check when implementing and installing. Otherwise, unexpected systems behaviors can arise that take time to chase down and resolve.

In this case, it was a confounding instance of I/O cards dropping off the local communications channel with no warning, error, or explanation.

SMOOTH INSTALL, TEST & STARTUP. THEN… NOT SO MUCH.

After install and testing the new CompactLogix system worked well and as expected. The remote I/O is mostly made up of 5069-OW16 relay output modules. There are also a couple of 5069-IB16 input modules mixed in.

After a day of production operations, the system began the unexpected behavior. Some parts of the machine would operate normally, while other parts did not operate at all. A quick inspection of the I/O rack showed blinking red “STATUS” lights on most, but not all I/O modules. The AENTR communications module reported no errors. The first three cards to the right of the comms module operated normally. The remaining cards in that rack did not show comms errors. They only showed “Not connected” in the card properties.

Cycling MOD power to the chassis would bring the system back to normal after the power up. The system would typically run for hours or another day before losing comms on the cards again.

INITIAL TROUBLESHOOTING

The first step for troubleshooting was to check firmware on the cards to current versions. Even though cards and firmware are expected to be backwards compatible, this was an easy solution to try. After updating project firmware to card revision levels, the communication drop behavior continued.

The next solution attempted was to move cards between slots in the I/O rack to see if the problem would follow the card. This did not improve on the problem, so that seemed to rule out hardware issues on the I/O comms backplane.

At this point it is important to note the OW16 relay cards in this system are in many cases driving AC inductive loads. Some loads are small interposing control relays, some are pneumatic solenoid controls, and others are motor contactors.

FORUMS TO THE RESCUE

Communication with Rockwell Automation support, along with initial online research, did not turn up problems or solutions that matched the behavior of this system. Additional tangential research late one night finally surfaced forum posts from 2018 describing the behavior found in this system.

It turns out that while SLC series 1746-OW16 may have some inherent resistance to inductive load noise, the 5069-OW16 cards do not. It appears the combination of the high speed data bus on the 5069 “backplane” and lack of internal surge suppression on the 5069-OW16 cards leave the system more susceptible to inductive noise than the system it is replacing. The existing system did not have surge suppression installed for the outputs.

INSTALLING SNUBBERS- PROBLEM RESOLVED

Since there were potentially dozens of output points to protect (not all outputs are being used), we searched for a cost-effective solution. We found RC snubbers made by Red Lion Controls (part number SNUB0000) and acquired enough to cover the project requirements.

The intention was to install a snubber at each used relay output point, but the panel wiring did not accommodate that type of installation. The fallback plan involved attempting to locate each of the inductive loads on the machine and in the control panel. When we could locate the load, the snubber was installed across the coil contacts of the load, which is best practice but not always feasible.

After installing the surge suppression at each load we could identify, the system has been running as originally intended, and with the upgraded control system. This is one instance where the system designer should take heed of the fine print on the 5069-OW16 indicating output snubbers are not integrated with card. For system stability, suppression of 120VAC inductive load transients is not optional.