Description
Hard-Numbers: Technical Specifications
- Processor: Intel 80386EX, 16 MHz clock
- User Program Memory: 40 KB (IC693CPU331)
- Register Memory: 240 KB (%R addressing)
- Floating Point: Supported (32-bit)
- Discrete I/O: 2048 points max combined (%I + %Q)
- Analog Input (%AI): 128 words (8K words with option modules)
- Analog Output (%AQ): 64 words (8K words with option modules)
- Internal Coils (%M): 1024 bits
- Discrete Global Memory (%G): 1280 bits
- Timers/Counters: 340 combined (up to 1280 with option)
- Scan Rate: 0.3 ms per 1K Boolean logic (typical)
- Serial Ports: 2 (Port 1: SNP/X master/slave, Port 2: SNP/X master/slave)
- Baud Rate: Up to 115.2 Kbaud
- Expansion: Yes (supports up to 7 baseplates including remote)
- Battery-Backed Clock: Yes (on-board battery)
- Power Draw: 950 mA @ +5 VDC
- Operating Temp: 0°C to 60°C (32°F to 140°F)
- Storage Temp: -40°C to 85°C (-40°F to 185°F)
- Module Type: Modular (plugs into CPU slot)
- Interrupts: Supported (up to 32 event interrupts)
- Subroutines: Supported (up to 64)
GE IC693CMM321
The Real-World Problem It Solves
You hit a wall at 320 I/O points and 12 KB of logic on the embedded CPUs. You’ve got a multi-stage process that needs expansion racks, analog loops requiring floating-point math, and actual time-of-day scheduling for shift changes. This CPU drops into a standard baseplate, gives you 2048 I/O capacity across seven racks, runs twice as fast, and has a clock that survives power outages.
Where you’ll typically find it:
- Batch process control: Pharmaceutical reactors, food/beverage mixing, chemical batch sequences—complex recipes with analog PID loops and time-based scheduling
- Large material handling: Airport baggage systems, distribution center sorters, automated storage/retrieval—multiple expansion racks coordinating hundreds of motors and sensors
- Assembly lines: Automotive body shops, appliance assembly—high-speed discrete control with coordinated motion and HMI integration
Bottom line: It’s the mid-range workhorse that bridges gap between small embedded CPUs and high-end 350/360 models—serious memory, expansion capability, and real-time clock without paying for features you won’t use.
Hardware Architecture & Under-the-Hood Logic
This is a modular CPU that plugs into any Series 90-30 baseplate CPU slot (leftmost position). The Intel 80386EX is a 32-bit processor running a 16-bit bus, giving you solid performance without the power draw of the 80486. Separate memory boards (optional) can be installed on the CPU module itself for additional register storage. Real-time clock circuitry lives right on the board with its own lithium battery.
-
Power-up sequence runs internal diagnostics on the 80386EX core, memory subsystem, and I/O interfaces. The CPU reads configuration data from non-volatile memory (NVRAM) and initializes the backplane communication driver. Battery-backed clock maintains time during power-off state.
-
Backplane communication occurs through a dedicated bus interface. The CPU scans all connected baseplates sequentially (Rack 0 through Rack X), gathering I/O images from each rack’s I/O modules. Expansion cables (IC693CBLxxx) carry data between CPU baseplate and remote racks.
-
Program scan executes ladder logic faster than embedded CPUs—0.3 ms per 1K Boolean. The 80386EX processes floating-point math in hardware, so PID loops and scaling operations don’t choke your scan time. Subroutines and program blocks allow modular programming structure.
-
Interrupts can break into the scan cycle for time-critical events. Up to 32 configurable interrupt inputs can trigger immediate subroutine execution. This is crucial for high-speed sensor response, emergency 终止 handling, or coordinating with motion controllers.
-
Serial port handlers for Port 1 and Port 2 operate independently. Each port can function as SNP or SNP-X master or slave. Master mode lets the CPU initiate communication to other PLCs, HMIs, or slave devices—something embedded CPUs can’t do. Baud rates up to 115.2 Kbaud allow faster data transfer.
-
Memory architecture splits between user program storage (40 KB) and register data storage (240 KB). The register memory (%R) is massive compared to embedded CPUs—critical for data logging, recipe storage, or large buffer areas for communication modules. Optional memory modules can expand register storage further.
-
Floating-point unit (FPU) on the 80386EX handles 32-bit IEEE 754 operations. You get real math, not integer scaling hacks. Temperature in °F, pressure in PSI, flow in GPM—keep your engineering units native throughout the logic. No more scaling by 100 and hoping you don’t overflow.
-
Battery-backed clock maintains year/month/day/hour/minute/second through power cycles. Lithium battery (CR2032 type typically) on the CPU board powers the clock circuit. When battery dies, you lose time-of-day but the CPU still runs—you just can’t schedule events based on actual time.
-
Power consumption runs higher than embedded CPUs—950 mA at +5VDC. That’s more than double the embedded CPUs. Check your power supply capacity (IC693PWR330 or PWR321 recommended) before loading up with high-current I/O modules. Add up all module currents plus 950 mA and verify you’re within supply rating.
-
Expansion capability allows up to 7 baseplates total: 1 CPU baseplate + 6 expansion or remote racks. Each rack requires its own power supply. CPU manages addressing across all racks automatically—no manual dipswitches. Addressing becomes %I + rack + module + point format automatically.
GE IC693CMM321
Field Service Pitfalls: What Rookies Get Wrong
Underestimating power supply load
You drop a 331 into a rack running IC693PWR321 and populate it with analog modules. The PLC resets randomly under load. The 331 draws 950mA alone—that power supply was already stressed with the old CPU and I/O.
- Field Rule: Do the math before you power up. IC693CPU331 = 950mA @ +5VDC minimum. Add every module’s current draw. If using PWR321 (output: 3A @ +5VDC), you’ve got roughly 2A left for everything else. PWR330 (5A @ +5VDC) is safer for loaded racks. Never exceed power supply rating—random faults and data corruption will haunt you.
Overlooking the clock battery
You’ve got the CPU installed for 5 years and time-based scheduling starts drifting. Shift changes happen at wrong times, datalog timestamps are worthless. The clock battery died, and nobody checked it.
- Field Rule: Replace the clock battery every 3-4 years during scheduled outages. It’s a CR2032 lithium cell on the CPU board. When battery voltage drops, you’ll get a low battery indicator bit (%S0012 typically). Don’t wait until time-of-day goes haywire—swap it proactively. The CPU keeps running with a dead battery, but scheduling functions become unreliable.
Assuming serial ports are SNP slave only
You try to set up peer-to-peer PLC communication and expect Port 1 to act like the embedded CPU slave mode only. It doesn’t work—you configured the wrong protocol for master mode.
- Field Rule: Both serial ports on the 331 can be SNP or SNP-X, master or slave. For PLC-to-PLC communication, you typically set one as SNP master initiating reads/writes to the other’s slave port. Check your programming software (Logicmaster, VersaPro, Proficy) serial port configuration carefully. Master mode requires explicit COMMREQ blocks in ladder logic—don’t assume it just works.
Mixing old and new expansion cables
You grab whatever expansion cable you find in the shop and string together racks. Communication between CPU and remote racks fails intermittently. Old cables aren’t impedance-matched or have pinout differences.
- Field Rule: Use the correct GE expansion cables for your rack spacing. IC693CBL7xx series for 10-foot, 50-foot, or 100-foot runs. Mismatched cables cause signal reflection, data corruption, and mysterious faults. Label your cables at both ends with length and part number. Don’t trust any cable that’s not marked—test it with a continuity checker or replace it.
Ignoring interrupt configuration
You need fast response to a sensor and set up an interrupt input. Nothing happens. You forgot to enable the interrupt in configuration and write the interrupt service routine subroutine.
- Field Rule: Interrupts require three things: (1) Configure the input as an interrupt source in hardware setup, (2) Write the interrupt service routine (ISR) as a subroutine, (3) Associate the ISR with the interrupt number in program configuration. The CPU won’t magically interrupt scan cycles just because you connected a wire. Test interrupts with a signal generator before trusting them in production.
Overrunning register memory with analog data
You install multiple high-density analog modules and suddenly the CPU faults for memory allocation. The %R memory is huge, but you didn’t account for module configuration consuming address space.
- Field Rule: Analog modules map to %R addresses. 64-word modules consume 64 registers each. Calculate your total analog register usage and verify it fits within 240 KB. If you’re running tight, use the AI/AQ configuration to disable unused channels—this frees up register space. Don’t assume “more is always better”—unused analog channels still eat memory.
Forgetting floating-point overhead
You rewrite all your old integer scaling logic to use floating-point math. Scan time doubles, and the CPU can’t complete the scan within watchdog timeout.
- Field Rule: Floating-point math is faster than integer emulation, but it’s still not free. Benchmark your scan time after converting to float math. If scan time approaches 200-300 ms, you’re flirting with watchdog timeout. Optimize critical rungs to use integer math where possible—reserve floating-point for actual analog loops and scaling operations.
Losing program during CPU swap due to dead battery
You pull the 331 for troubleshooting and set it on the shelf for a week. When you reinstall, the program’s gone. The NVRAM battery was weak, and without power, it lost everything.
- Field Rule: Back up your program before removing any CPU from service. NVRAM retention depends on battery condition—old CPUs may only hold programs for days without power. Use the serial port or PCM card to download a backup. Always assume a powered-down CPU will lose its program unless you’ve verified battery health recently.
Misconfiguring rack addressing
You install expansion racks and expect the CPU to automatically find I/O points beyond the first rack. Nothing works in Rack 2. You forgot to configure the rack-to-rack addressing in the CPU configuration.
- Field Rule: The CPU needs to know what’s in each rack. In programming software, define each rack (0-7) and what modules are in each slot. The CPU uses this configuration to build the I/O image tables. If configuration doesn’t match physical hardware, you’ll get mismatch faults or I/O that simply doesn’t respond. Always re-verify configuration after adding or moving racks.
Swapping CPU modules while power is applied
You hot-swap a 331 to avoid a production 终止page. You blow the backplane driver IC on the new CPU. Backplane transients kill module electronics instantly.
- Field Rule: NEVER hot-swap modular CPUs. Power down the entire rack before removing or installing any CPU. Backplane voltage spikes when modules are inserted/removed under load. If you need redundancy, design a hot-standby CPU system with proper transfer switch—don’t jury-rig hot swaps. A blown CPU costs a lot more than a planned shutdown.
Commercial Availability & Pricing Note
Please note: The listed price is for reference only and is not binding. Final pricing and terms are subject to negotiation based on current market conditions and availability.
