Maintaining Your GE RX3i System with IC695CPU310 Health Checks

In the world of industrial automation, equipment reliability is the only metric that truly counts. While new controllers hit the market every year, the GE Fanuc (now Emerson) RX3i series remains a workhorse in factories across the globe. At the center of many of these systems is the IC695CPU310, a 300 MHz processor that bridges the gap between older serial-based setups and modern high-speed architectures.

Keeping these units running requires more than just hoping for the best. The CPU310 has specific needs regarding memory retention, heat management, and power delivery that differ from newer models. If you treat it exactly like a modern Ethernet-based CPU, you might run into unexpected downtime. This guide breaks down the specific health checks you need to perform to ensure your processing lines stay active and your code remains safe.

Understanding What Makes Your CPU Tick

To maintain this processor effectively, you have to understand its internal limitations and how it handles data compared to other units in the family. The CPU310 is unique because it manages two different backplane buses simultaneously, which creates a specific set of maintenance priorities.

The 300 MHz Processing Limit

The IC695CPU310 operates at 300 MHz. While this is sufficient for many applications, it is a finite resource. Over years of operation, maintenance teams often add new lines of code, additional PID loops, or more complex motion instructions to the program. This "code bloat" gradually increases the sweep time—the time it takes the processor to read inputs, solve logic, and write outputs.

A critical health check involves monitoring this sweep time during peak operation. If your sweep time creeps up and gets too close to the watchdog timer limit (usually set to 200ms), the CPU will trip and stop the machine to prevent unsafe operation. Unlike faster 1 GHz processors, this unit has less headroom for inefficient code.

The Dual-Bus Balancing Act

One reason facilities migrated from the legacy IC693CPU364 to the RX3i platform was the ability to mix old and new technology. The CPU310 manages a high-speed PCI bus for modern modules and a traditional serial bus for older Series 90-30 style cards.

This dual nature means you have two "health zones" to watch. A failure in a legacy serial module can bog down the CPU as it tries to manage the error, which indirectly slows down your high-speed PCI modules. When troubleshooting, remember that a slowdown in your high-speed data might actually be caused by a loose connection on a vintage serial card sharing the same rack.

The CPU310 is a robust bridge between old and new technologies, but its processing speed is finite. Regular monitoring of scan times and bus errors ensures that legacy hardware issues do not impact overall system performance.

Physical Health and Power Checks

Electronics degrade over time due to heat and electrical stress, and the RX3i rack is no exception. A physical inspection is the cheapest and most effective way to prevent hardware failure.

Verify the 4-Inch Clearance Rule

The CPU generates heat that must dissipate through natural convection. The manual specifies a minimum clearance of four inches (102mm) on all sides of the rack. If maintenance teams have installed wire ducts or other devices too close to the vents, heat will build up. This can cook the electrolytic capacitors inside the IC695PSA040 power supply, leading to unstable voltage on the backplane. Check that the airflow path is completely unobstructed.

Inspect Grounding and Shielding

Electrical noise can cause "ghost" faults where the CPU thinks a module has failed. Check the grounding bar at the bottom of the rack. Ensure the screws are tight and that cable shields for analog modules are grounded at the rack end. Loose grounds are a common cause of intermittent data corruption.

Check Module Seating

Vibration from heavy machinery can cause modules to "walk" out of their slots slightly. This is especially true for the heavy CPU and power supply modules. Verify that the plastic latch at the bottom of every module is fully engaged and clicked into place. A module that is slightly loose will cause random "Loss of I/O" faults.

Clean the Backplane

In dusty environments like foundries or paper mills, conductive dust can settle on the backplane pins. If you are doing a shutdown maintenance, use dry compressed air to blow out the backplane. Never use solvents. Dust accumulation can bridge the fine-pitch pins of the PCI bus, causing short circuits.

Heat and vibration are the primary enemies of the RX3i hardware. Ensuring proper ventilation space and tight physical connections will prevent the majority of intermittent electrical faults.

The Critical Battery Mistake to Avoid

The most distinct maintenance requirement of the CPU310 is its reliance on a battery for memory retention. Unlike newer units that use capacitors to write data to flash memory upon power loss, the CPU310 needs a healthy battery to keep your program alive in RAM when the power is off.

Using the correct battery pack and replacing it while the system is powered on is non-negotiable for this processor. Failing to do so guarantees data loss during the next power outage.

Reading the Lights on the Faceplate

The LED faceplate provides an immediate health report without needing to connect a laptop. Understanding the blink codes can save you an hour of troubleshooting time.

The Standard "Go" Signals

In a healthy system, you want to see the OK, RUN, and OUTPUTS ENABLED lights showing solid green. The BATTERY and SYSTEM FAULT lights should be off. If the COM lights are blinking, that just means data is moving through the serial ports, which is normal if you have an HMI connected.

The "Blinking Unison" Error

There is a specific error pattern that terrifies technicians: the RUN and OUTPUTS ENABLED lights blinking together in unison (sometimes with other LEDs). This does not usually mean the hardware is fried. It typically means the CPU is stuck in Boot Mode. This happens if a firmware update was interrupted or the operating system corrupted. You can often recover from this by using the WinLoader utility to reload the firmware, rather than scrapping the unit.

The Battery Warning

If the BATTERY LED is blinking red, you have a limited window—often just days or weeks—before the battery voltage drops below the level required to retain memory. Do not ignore this light. If the light is solid red, the battery is likely dead or disconnected.

The LEDs are your first line of defense in diagnostics. Learn to distinguish between a simple battery warning and a boot mode error to react appropriately to downtime events.

Software Diagnostics and Advanced Modules

While physical checks are vital, the most detailed health information comes from the software side. You need to interface with the controller using Proficy Machine Edition (PME) to see what is happening under the hood.

Analyze the Controller Fault Table

The CPU logs internal errors here. Look for "Group 140" errors, which are non-critical events, versus "Group 1" errors, which are fatal. If you see repeated "Loss of Module" errors, it usually points to a loose module or a failing backplane slot.

Manage High-Availability Modules

If your system requires redundancy, you might be using an IC695RMX128 synchronization module. This module uses fiber optics to link two CPUs. A standard health check involves verifying the fiber signal strength. Dirty fiber tips are a common failure point. Ensure the fiber cables have not been bent beyond their bend radius, as this attenuates the signal and causes the redundancy link to drop.

Use WinLoader for Firmware

Updating firmware on the CPU310 is different from newer Ethernet CPUs. You must use a serial cable and a utility called WinLoader. A common trick: if the update fails, try removing other I/O modules from the rack. The power cycling process during the update can sometimes be disrupted by the startup load of other modules.

Clear Faults After Repair

The system will often keep the "System Fault" LED lit even after you fix the problem. You must manually clear the Controller and I/O Fault tables using PME. This is also a good time to save the fault log to your PC for historical tracking before you wipe it.

Regularly connecting to the processor to review and clear fault tables allows you to spot intermittent issues before they become hard failures. Software tools provide the granular detail needed for true reliability.

Conclusion

The IC695CPU310 is a capable processor, but it demands a proactive maintenance strategy. By respecting its thermal limits, managing the specific battery requirements, and utilizing the diagnostic tables, you can extend the life of your control system significantly. Do not wait for the red light to turn on. Schedule your battery replacements and backplane inspections today to keep your facility running smoothly.

FAQs

Q1: Can I use the IC690ACC001 coin cell battery in my IC695CPU310?

No, you cannot. The IC695CPU310 requires the IC698ACC701 lithium battery pack to maintain its 10 MB of user memory. The coin cell battery (IC690ACC001) is designed for different CPU models like the CPE305 or CPE330 specifically for the Real-Time Clock and lacks the capacity and connection type to support the CPU310's RAM.

Q2: What does it mean when the RUN and OUTPUTS ENABLED lights blink in unison?

This blink code indicates the CPU is in Boot Mode or firmware update mode. It typically occurs after a failed firmware update or a corrupted operating system load. The CPU is waiting for a valid firmware file. You can usually resolve this by reconnecting the WinLoader utility and restarting the firmware upgrade process.

Q3: How do I clear the "System Fault" light on the CPU?

You must use the Proficy Machine Edition programming software. Connect to the PLC, go online, and ensure you are in "Programmer Mode." Navigate to the Diagnostics menu, open the Controller Fault Table and I/O Fault Table, and click the "Clear Faults" button. Cycling the power will not clear these historical fault logs.

Q4: Is it safe to replace the battery while the PLC is powered on?

Yes, replacing the battery while the system is powered ON is the recommended procedure. If you remove the battery while the power is off, the volatile memory (RAM) will lose power, and your program and data will be erased immediately. Changing it under power ensures the program is retained.