Description
Key Technical Specifications
|
Technical Category
|
Specifications
|
|---|---|
|
Model Identification
|
ABB RET670 1MRK002816-AC
|
|
Manufacturer
|
ABB Power Grids Division
|
|
Protection Functions
|
Differential protection (ANSI 87T), overcurrent (50/51), overvoltage (59), undervoltage (27), temperature monitoring (49T)
|
|
Performance
|
5ms fault detection latency, 0.1% current measurement accuracy
|
|
Communication
|
IEC 61850 (GOOSE/MMS), IEC 60870-5-103, DNP3.0, Ethernet (100Mbps)
|
|
Redundancy
|
Dual power supply (24/48/110/220V DC/AC), hot-standby configuration support
|
|
Operating Environment
|
-40°C to +70°C (-40°F to +158°F), 5-95% humidity (non-condensing)
|
|
Certifications
|
IEC 61508 (SIL 4), IEC 60255, CE, UL 61850, CCC
|
|
Mechanical Design
|
Panel mount, IP42 front panel, 19” rack compatible
|
ABB RET670 1MRK002816-AC
Field Application & Problem Solved
In 220kV-750kV substations, transformer protection errors directly lead to cascading outages. A 2024 outage at a North China 500kV substation cost $2.3 million when a legacy relay misidentified transformer inrush current as a short circuit, tripping a 315MVA transformer and cutting power to 120,000 residents. Older relays also lack digital integration: a Southeast Asian substation spent 4 hours manually collecting fault data after a transformer failure, delaying restoration by 2 hours.
The RET670 is the “safety gate” for power transformers, with core value in accurate fault discrimination + digital efficiency. Its adaptive differential protection algorithm solved the North China substation’s issue: by analyzing current waveform characteristics, it distinguishes inrush current (non-fault) from short-circuit current with 99.9% accuracy, eliminating false trips. At a Texas 345kV substation, the IEC 61850 GOOSE function transmitted fault data to the control center in 300ms, cutting troubleshooting time from 4 hours to 20 minutes.
SIL 4 certification is mandatory for critical substations. A European substation avoided $800k in non-compliance fines by replacing non-certified relays with RET670. In a Saudi Arabian 750kV substation, the dual power supply design prevented an outage during a DC system failure—one power input switched to standby in 2ms, keeping protection active. The relay’s temperature monitoring function also detected a winding overheating issue at a Brazilian substation 3 days before a potential breakdown, avoiding $1.5 million in transformer repair costs.
The PFPL203 is the “power backbone” of AC800M racks, designed to eliminate single points of failure. Its core value is redundant reliability plus energy efficiency. The 1+1 redundancy fixed the Gulf Coast refinery’s issue: when a lightning strike damaged one PFPL203, the redundant unit took over seamlessly—no voltage dip, no process upset. At the Pennsylvania plant, the 92% efficiency cut power consumption by 1.4kW/day, saving $600/year per module (they used 12 units, totaling $7.2k/year in savings).
Its hot-pluggable design is critical for scheduled maintenance. A Texas LNG plant swapped a faulty PFPL203 in 3 minutes (vs. 20 minutes with non-pluggable units), keeping their liquefaction train online. In corrosive offshore platforms, the module’s conformal-coated PCB outlasted legacy PSUs by 2x—critical when replacing a module at sea costs $40k in vessel time. The built-in overvoltage protection also prevented a $20k CPU damage at a Florida pharmaceutical plant when the mains voltage spiked to 260V AC.
The GVC736BE101 is the “fuel regulator” for gas turbines, acting as the link between the Symphony Plus DCS and the turbine’s fuel valves. Its core value is precision plus fail-safe redundancy. The ±0.1% control precision fixed the European plant’s fuel waste issue, cutting consumption by 1.8% and saving $270k/year. At a Texas GT36 plant, the 1+1 redundancy eliminated unplanned outages: when a lightning strike damaged the primary controller, the standby took over in 4ms—no flameout, no downtime.
Its SIL 3 certification is non-negotiable for safety-critical applications. A Middle East plant replaced non-certified controllers with GVC736BE101 to meet local regulatory requirements, avoiding $150k/quarter in non-compliance penalties. The IEC 61850 integration let technicians monitor valve position and fuel flow in real time via the DCS, reducing troubleshooting time for minor drift issues from 4 hours to 20 minutes.
The DSQC346E is the “signal bridge” between ABB robots and factory equipment—sensors, grippers, conveyors, and safety interlocks. Its core value is speed plus ruggedness for robot cycles. The 1ms response time fixed the Texas plant’s component miss issue, cutting scrap to 0.1%. At a Ohio welding facility, its 2A output channels directly power small weld guns (no external relays needed), reducing wiring by 30% and eliminating relay failure points that caused 2 hours of downtime weekly.
Its hot-pluggable design is a game-changer for scheduled maintenance. A Mexico automotive plant swaps faulty modules in 5 minutes (vs. 30 minutes with non-pluggable units), keeping their 24/7 production line running. In dusty foundry environments, the module’s conformal-coated PCB resists metal dust buildup—unlike generic modules that short out after 6 months, the DSQC346E runs reliably for 3+ years.
In AC800M DCS installations—especially in cramped refinery control cabinets—poor terminal management causes more unplanned downtime than most engineers realize. A 2023 outage at a Midwest chemical plant cost $320,000 when a loose wire on a generic terminal block caused a pH sensor signal to drop, triggering an emergency shutdown of a reactor. Legacy terminals also struggle with signal interference: at a Texas refinery, unshielded terminals let motor noise corrupt temperature readings, leading to 3% off-spec fuel and $150k/week in losses. Another headache: traditional screw terminals require torque wrenches and take 2x longer to wire, delaying commissioning by days.
The 2GUS031803A1 is the unsung hero of clean, reliable I/O wiring in AC800M racks. It’s not just a terminal block—it’s a signal integrity hub. Its core value is fast, interference-free wiring that stays tight. The tool-less spring terminals cut wiring time by 60% at a Louisiana LNG plant, getting their new liquefaction train online 3 days early. The integrated shield clamp (no extra hardware needed) eliminated the refinery’s noise issue—temperature readings stabilized within 0.1°C, and off-spec product dropped to 0.2%.
Its universal design is a lifesaver during retrofits. A California power plant used it to replace 4 different terminal types across their AC800M system, standardizing parts and cutting spare inventory costs by $18k. When a sensor failed, technicians swapped wires in 2 minutes (vs. 15 with old terminals) because the color-coded channels and clear labeling eliminated guesswork. In corrosive environments, the nickel-plated terminals resist oxidation—something generic brass terminals can’t match, as seen at a Florida wastewater plant where legacy terminals corroded and failed after 18 months, while the 2GUS031803A1 units are still going strong after 3 years.
Field Application & Problem Solved
In power generation and petrochemicals—think a 600MW coal-fired plant’s boiler control or a refinery’s FCC unit—controller failure isn’t just downtime; it’s a safety hazard. A 2023 outage at a North China power plant cost $1.8 million in 8 hours when its legacy single-CPU controller crashed during a grid fluctuation, triggering a boiler trip. Even with basic redundancy, older controllers struggle with two fatal flaws: failover times over 50ms that cause process upsets, and scan times of 8ms that make turbine speed control unstable—leading to costly overspeed trips.
The PPC905AE101 is the “safety net” in Symphony Plus DCS racks for these high-stakes scenarios. Its core value is instant failover plus pinpoint control. The 1+1 redundancy fixed the North China plant’s issue: when a voltage spike damaged the primary CPU, the standby took over in 4ms—fast enough that the boiler’s water level never deviated from setpoint. At a Texas gas turbine plant, the 1ms scan time cut speed control deviation from ±0.5Hz to ±0.1Hz, eliminating 3 annual overspeed trips that each cost $250k.
Its IEC 61850 GOOSE support is transformative for smart power plants. A California combined-cycle plant used it to replace 20-year-old controllers, integrating directly with digital turbine governors and smart circuit breakers. This eliminated 8 legacy gateways, cutting communication latency by 90% and letting technicians troubleshoot via real-time data streams instead of manual gauge checks. In corrosive offshore platforms, its conformal-coated PCB outlasted older controllers by 3x—critical when replacing a controller at sea costs $50k in vessel time.
You’ll find the PM864 as the “brain” of AC800M DCS racks in these high-stakes environments. It’s not just a CPU—it’s a fail-safe control hub. Its core value is zero unplanned downtime plus speed. The 1+1 redundancy fixed the Gulf Coast refinery’s issue: when a lightning strike fried the primary CPU, the standby took over in 8ms—no process upset, no downtime. The dual-core 800MHz processor cut scan times from 10ms to 2ms at a Pennsylvania chemical plant, stabilizing their reactor temperature control and reducing off-spec product by 90%.
Its IEC 61850 compliance is a game-changer for modernization. A California power plant used it to replace 15-year-old CPUs, integrating directly with new smart breakers and turbines without gateways. This cut communication faults by 100% and simplified troubleshooting—technicians now pull real-time data from instruments via the CPU, not via 3rd-party software.
ABB RET670 1MRK002816-AC
Installation & Maintenance Pitfalls (Expert Tips)
Redundant Wiring—Must Connect to Separate AC Sources
Plugging two PFPL203s into one AC circuit defeats redundancy. A Ohio power plant did this; a circuit breaker trip took down both modules, triggering a 2-hour DCS shutdown. Feed Module A from the plant’s main AC bus, Module B from the emergency AC bus. Use ABB’s power monitoring relay (1SVR730700R0000) to alarm if either AC input drops.
Load Sharing—Don’t Mix PFPL203 with Older PFPL202 Modules
Mixing PFPL203 (10A) with PFPL202 (5A) in a redundant pair causes uneven load distribution. A Louisiana chemical plant did this, and the PFPL202 overheated (100% load vs. PFPL203’s 30%). Stick to identical models for redundancy—use the module’s LED status (green = normal, amber = half load) to verify balanced sharing.
Ventilation—Leave 50mm Clearance Above/Below
Blocking airflow around the PFPL203’s top/bottom vents causes overtemperature shutdown. A Wyoming coal plant stacked IO modules directly above the PSU; during summer, the PFPL203 shut down at 2pm daily (cabinet temp 55°C). Moving the modules to create 50mm clearance eliminated the issue—never use cable ties across the vent slots.
定期负载测试—Every 6 Months (Not Annual)
Waiting for a failure to test the PSU hides degradation. The PFPL203’s output current degrades over time; a California refinery’s module dropped to 7A (from 10A) after 18 months, causing CPU brownouts. Use ABB’s load tester (3BSE048901R1) to apply 10A load for 5 minutes—replace if output voltage drops below 23.5V DC.
Calibration—Do It Every 6 Months (Not Annual)
Relying on annual calibration leads to hidden drift. The GVC736BE101’s precision degrades over time due to temperature cycles; a Chinese plant skipped 6-month checks, and valve position drifted by 0.3%—triggering a turbine load reduction. Use ABB’s Calibration Toolkit (3BSE050100R1) with a certified test stand to verify linearity. Always calibrate both controllers in the redundant pair to ensure matching setpoints.
Rookies plug the module in without assigning a unique address, causing it to clash with existing I/O. A Indiana robot cell had two DSQC346E modules fighting for the same address, leading to random gripper failures. Use ABB’s RobotStudio software to set a unique RCI bus address (1-16) before powering on—label the module with the address to avoid mix-ups during swaps.
Redundant Power—Feed from Separate Source
Plugging both controller power inputs into one supply defeats redundancy. A German plant did this; a power dip took down both units, causing a 2-hour turbine shutdown. Feed Input A from the turbine’s main UPS, Input B from the plant’s emergency UPS. Use the controller’s built-in power monitor to alarm if either input drops below 20VDC.
The 16 output channels are rated 2A max—connecting a 3A gripper motor directly fries the channel. A Tennessee packaging plant did this, costing $300 in module repairs and 2 hours of downtime. Use ABB’s compact contactors (1SBL161001R8001) for loads >2A, and add a 2A fuse per channel (ABB part 1SFA616001R1000) as a safety net.
Wiring Polarity—Reverse Polarity Fries Actuator Output
Swapping positive/negative on the 4-20mA actuator output damages the controller’s DAC. A Texas plant made this mistake during retrofitting, costing $2k in repairs and 4 hours of downtime. Use color-coded wiring (red = +, black = -) and verify with a multimeter before power-up—label terminals clearly with ABB’s heat-shrink markers (3BSE013251R1).
Daisy-chaining I/O cable shields causes noise in sensor signals. A California battery assembly robot kept false-triggering safety stops until we grounded each DSQC346E’s shield clamp to the robot chassis (not the cabinet ground). Use 360° shield clamps (1SFL500001R0001) to ensure solid contact—loose shields = erratic robot behavior.
Environmental Sealing—Use Gland Plates in Humid Areas
Unsealed cable entries let moisture enter the controller. A Florida plant’s GVC736BE101 failed after 18 months due to tropical humidity. Install ABB’s IP67 cable glands (3BSE048900R1) and use dielectric grease on connector pins—this extends service life to 5+ years in harsh environments.
Partial backplane insertion leads to intermittent communication drops. A Michigan welding robot randomly lost I/O signals; the issue was a DSQC346E not fully latched into the IRC5 cabinet. Push firmly until the blue latch clicks—verify with RobotStudio’s “I/O Module Status” tool to confirm a solid connection.
Don’t Skip the Shield Clamp—Grounding Isn’t Optional
Rookies wire the signal but ignore the shield, turning the terminal block into an antenna for electrical noise. The 2GUS031803A1 has a built-in shield clamp (green terminal on the left) that must be connected to the cabinet ground bar with 4mm² wire. A Pennsylvania refinery skipped this, and their flow meter signals fluctuated by 10%—enough to trigger process alarms. Crimping the shield to a ferrule and tightening it to the clamp fixed the issue in 10 minutes. Never daisy-chain shields; each terminal block’s clamp needs a dedicated ground path.
Spring Terminals—Don’t Overforce the Wire
Tool-less terminals are fast, but shoving 3mm² wire (exceeding the 2.5mm² max) into the 2GUS031803A1’s channels bends the springs, causing loose connections. A Wyoming coal plant did this with 4-20mA current loops; 2 weeks later, 6 terminals failed, leading to a conveyor shutdown. Use wire strippers with a 6mm strip length (match the terminal’s marking) and confirm the wire clicks into place—tug gently to ensure it doesn’t pull out. For stranded wire, always use a ferrule (ABB part 1SFA896101R1000) to prevent fraying.
Terminal Grouping—Separate Power and Signal Wires
The 16 channels are split into 4 groups of 4—mixing 240V AC power wires with 4-20mA signal wires in the same group causes cross-talk. A Ohio pharmaceutical plant made this mistake, and their pressure sensor signals drifted by 0.5psi. Reorganize wires: group power in channels 1-4, analog signals in 5-12, and digital in 13-16. Use the module’s built-in dividers to separate groups—they’re not just for show. Label each channel with ABB’s heat-shrink markers (3BSE013251R1) to avoid mix-ups during troubleshooting.
Corrosion Prevention—Use Dielectric Grease in Wet Environments
In humid or coastal locations (e.g., Florida, Texas Gulf Coast), moisture creeps into terminals and causes corrosion. A wastewater plant in Tampa saw green oxidation on terminals after 6 months, leading to high resistance and signal loss. The fix: apply a thin layer of dielectric grease (ABB recommended: 3BSE013252R1) to the wire strands before inserting them into the terminal. This creates a moisture barrier—we checked the terminals 2 years later, and they were still clean.
Sync Link—Fiber Optic Only for Long Distances (No Copper)
Rookies use copper Cat6 for redundant sync links over 10 meters—electrical noise corrupts data. The PPC905AE101 requires ABB’s 50/125μm multimode fiber (part 3BSE048899R1) for distances >10m; copper only works for rack-mounted pairs. A Louisiana refinery used 20m copper links, causing “sync loss” alarms every time a nearby motor started. Switching to fiber eliminated the issue—fiber’s EMI immunity is non-negotiable for plant floors.
Dual Power Inputs—Must Feed from Separate UPS Sources
Plugging both CPU power inputs into one UPS defeats redundancy. A Pennsylvania coal plant did this; a UPS failure took down both CPUs, triggering a 3-hour boiler shutdown. The fix: feed Input A from the plant’s main UPS, Input B from the emergency UPS. Use ABB’s power monitoring module (CI867K) to alarm if either input drops—don’t rely on the controller’s built-in LED alone.
Firmware Update—Never Skip the “Sync Config” Step
Updating one CPU’s firmware without syncing the config erases critical redundancy settings. A Dubai desalination plant updated the primary CPU to v6.3 but forgot to sync; when it failed, the standby rejected the handoff because of mismatched turbine control logic. Always use ABB’s Engineer IT software to run “Redundancy Sync” after firmware updates—this copies not just firmware, but all control loops and I/O mappings.
I/O Module Termination—Torque to 0.8N·m (Not Hand-Tight)
Loose I/O terminal screws cause intermittent signal loss. The PPC905AE101’s compatible AI835A modules require 0.8N·m torque—hand-tight is usually 0.3-0.5N·m. A Germany chemical plant’s reactor temp sensors flickered until we re-torqued all terminals with ABB’s torque screwdriver (3HNA010777-001). Label terminals with torque values—technicians won’t guess.
Rookies use Cat5e Ethernet cable for the CPU sync link—big mistake. The PM864’s sync requires shielded, twisted-pair cable (ABB part 3BSE013249R1) to avoid noise interference. A Louisiana chemical plant used generic Cat5e; during a motor startup, electrical noise corrupted the sync signal, causing both CPUs to go “active” (split brain) and trip the process. Replacing with ABB’s cable fixed it.
Firmware Must Match Across Redundant Pair & I/O Modules
Mismatched firmware (e.g., CPU A v5.1, CPU B v5.0) causes failover failures. A Texas LNG plant updated one CPU but not the other—when the primary failed, the standby rejected the handoff, triggering a 4-hour shutdown. Always use ABB’s System 800xA Engineering Studio to flash both CPUs simultaneously, and verify I/O module firmware (e.g., AI810) is within 1 version of the CPU.
Battery Backup—Replace Every 2 Years (Not When It Dies)
The PM864’s RAM battery (3BSE030220R1) keeps process data during power outages. Waiting for the “low battery” alarm is risky—batteries fail suddenly in extreme heat. A Wyoming refinery ignored the alarm; a power dip erased 2 hours of batch data, costing $30k in rework. Schedule annual inspections, replace every 2 years, and test after replacement by pulling the power.
Grounding—Dedicated Path to Plant Ground Grid
Daisy-chaining the CPU’s ground to I/O modules causes noise in analog signals. A Florida pharmaceutical plant’s pH sensor readings fluctuated by 0.5pH until we ran a 6mm² copper wire from the CPU chassis directly to the plant’s ground grid (not the cabinet ground). Use 360° shield clamps to ensure solid contact—loose grounds = unstable control.
The DSQC346E is a dedicated digital I/O module for ABB’s IRC5 robot controllers, engineered to handle the fast, repetitive signals of industrial robot cycles. At its core, a 16-bit microcontroller processes input signals and triggers outputs, communicating with the robot’s main CPU via the RCI bus—this offloads I/O processing from the robot’s core controller, keeping cycle times consistent.
Its 16 input channels use sinking logic (common positive) to interface with proximity sensors and limit switches, while the 16 sourcing outputs drive grippers, solenoids, and indicator lights. Each output has built-in short-circuit protection: if a wire shorts to ground, the module shuts down the channel (no damage) and logs a fault in RobotStudio for quick troubleshooting.
Ruggedization is tailored for factory floors: the ABS plastic housing resists oil and coolants, and the 0°C to +45°C operating range handles temperature swings in welding bays. The hot-pluggable design uses spring-loaded backplane connectors that maintain power during replacement—critical for 24/7 production. It’s not just an I/O module; it’s a reliable link that keeps robots in sync with factory equipment, where even 1ms of latency can break a production cycle.
The 2GUS031803A1 is a universal I/O terminal module engineered to solve the messy, error-prone wiring challenges of AC800M DCS systems. At its core, it’s a high-density signal distribution hub—16 channels in a 45mm-wide package—that accepts nearly every signal type used in industrial control, from low-voltage digital inputs to thermocouple signals and 4-20mA analog loops. The tool-less spring terminals are the key innovation: they use a cage-clamp design that grips wires firmly without torque wrenches, ensuring consistent contact pressure (1.5N typical) that resists vibration—critical in pump rooms and turbine halls.
Signal integrity is built into every detail. The nickel-plated copper terminals resist corrosion and minimize contact resistance (≤5mΩ), while the 600V AC isolation between terminal groups prevents ground loops and cross-talk. The integrated shield clamp provides 360° contact with cable shields, diverting electromagnetic interference (EMI) from nearby motors or VFDs away from sensitive signals. Unlike generic terminals, the 2GUS031803A1 is designed to mount directly onto AC800M I/O racks (ER801/ER802) using ABB’s quick-mount brackets, eliminating the need for extra DIN rail space.
Ruggedization makes it suitable for harsh environments: the polyamide housing resists oil, chemicals, and UV exposure, and the -40°C to +85°C operating range handles everything from arctic pipeline control cabinets to desert power plant enclosures. In critical applications, the module’s clear labeling and color-coded channels reduce human error during wiring and troubleshooting—something that matters when a single wrong wire can cost six figures in downtime. It’s not the flashiest DCS component, but it’s the one that keeps signals reliable, wiring clean, and technicians from pulling their hair out during commissioning.
Installation & Maintenance Pitfalls (Expert Tips)
CT/PT Wiring—Polarity Error Causes Protection Failure
Reversing current transformer (CT) polarity wires turns differential protection into a “blind spot.” A Indian 220kV substation did this during commissioning; the RET670 failed to detect a phase-to-phase short circuit, leading to 10 minutes of transformer overload. Use ABB’s CT Polarity Tester (1MRK000900R0001) to verify: the “P1” terminal of CT must connect to the relay’s “I1” input. Mark wires with phase labels (A/B/C) and check the relay’s “Current Balance” menu to confirm differential current <0.1A.
Setting Configuration—Don’t Ignore Inrush Blocking Parameters
Default inrush blocking settings don’t fit all transformers. A Canadian substation used factory settings for a 500MVA transformer, causing 2 false trips during energization. Use ABB’s Relay Configurator (PCM600 v4.10) to set “Inrush Blocking Time” based on transformer capacity: 100-300MVA = 200ms, 300-750MVA = 300ms. Always test energization 3 times after configuration to validate blocking effectiveness.
Communication—GOOSE Time Synchronization Is Mandatory
Unsynchronized GOOSE messages cause protection coordination errors. A German substation’s RET670 and circuit breaker relay had a 50ms time difference, leading to 200ms delay in tripping. Deploy IEEE 1588 PTP (Precision Time Protocol) with a time server (ABB SPC600) to synchronize all devices within ±1ms. Check the relay’s “PTP Status” page to confirm synchronization state—”Locked” indicates normal operation.
Regular Testing—Annual Differential Calibration Can’t Be Skipped
Long-term operation causes CT ratio drift. A Australian substation skipped annual calibration; the RET670’s differential current deviated to 0.5A, increasing false trip risk. Use ABB’s Test Set (1MRK001000R0001) to inject 5A, 10A, 20A test currents into each phase, verifying the relay’s measured value error <0.1%. Record calibration data in PCM600 to track drift trends.
Technical Deep Dive & Overview
The RET670 1MRK002816-AC is a flagship digital protection relay in ABB’s Relion® series, dedicated to power transformer protection. Its core is a dual-core 32-bit MCU architecture: one core runs protection algorithms (differential, overcurrent) with 1ms scan cycle, while the other handles communication and data processing, avoiding algorithm delays caused by communication bottlenecks.
The adaptive differential protection algorithm is its technical breakthrough. Traditional relays use fixed threshold settings, which are prone to misjudgment when facing inrush current or CT saturation. The RET670 analyzes 8 characteristic parameters of current waveforms (peak factor, harmonic content, etc.) in real time, establishing a dynamic threshold that changes with the operating state. For example, during transformer energization, the threshold automatically rises to block inrush current; when a real short circuit occurs, the threshold drops rapidly to ensure fast tripping.
Digital integration is fully compliant with IEC 61850 standards. The GOOSE message transmission latency is <1ms, enabling “trip-by-message” without hard wiring, reducing cable usage by 60% in substations. The MMS protocol supports remote configuration and fault recording—technicians can modify protection settings or extract 64-point fault wave records from the control center, eliminating on-site operations. The relay’s 16GB flash memory can store 10,000 fault records, providing data support for accident analysis.
Ruggedization adapts to harsh substation environments: the conformal-coated PCB resists salt spray and dust (suitable for coastal substations), and the -40°C to +70°C operating range handles extreme cold in northern regions and high temperature in desert areas. SIL 4 certification ensures that even in single-fault conditions, the relay can still perform protection functions, meeting the highest safety requirements of power grids. For substations where transformer failure costs $10k per minute, the RET670 is not just a relay—it’s a cost-saving and safety-guaranteeing core device.
Redundancy is its defining feature. When two PFPL203s are configured in parallel, a built-in redundancy controller enables automatic load sharing (each module carries 50% of the load at full 10A output). If one module fails (e.g., AC input loss, overtemperature), the other ramps to 100% load in <10ms—faster than the AC800M’s voltage dropout tolerance (20ms), so no CPU or IO shutdown. The module’s hot-pluggable design uses spring-loaded connectors that maintain output power during replacement, critical for 24/7 processes.
Protection mechanisms are layered for industrial robustness. Overvoltage protection cuts output at 28V DC to prevent CPU/IO damage from voltage spikes; overcurrent protection trips at 16A to avoid wire overheating; overtemperature shutdown activates at 75°C, with a pre-alarm (amber LED) at 70°C. The 3kV AC input-output isolation blocks ground loops and electromagnetic interference (EMI) from nearby motors or VFDs—critical in refineries with high EMI environments.
Ruggedization suits harsh industrial settings: the aluminum heatsink dissipates heat without fans (no moving parts to fail), and the -10°C to +60°C operating range handles unconditioned control cabinets in desert or arctic locations. The compact 17.5mm width lets two redundant modules fit in 35mm DIN rail space, saving cabinet real estate. This isn’t just a power supply—it’s the unsung hero that keeps AC800M DCS systems running, where a single power glitch can cost millions in downtime.
The 3BHE014070R0101 PPC905AE101 is the flagship redundant controller for ABB’s Symphony Plus DCS, engineered to eliminate single points of failure in critical control. At its core, dual Intel Atom E3827 CPUs operate in parallel: one runs control logic (turbine speed loops, boiler level control) while the other manages I/O data and communication—eliminating the “communication bottleneck” of single-core designs. The 4GB ECC RAM corrects memory errors (common in high-EMI environments), and 16GB flash stores firmware and application code with wear-leveling for 10+ year longevity.
Redundancy is its crown jewel. The primary CPU sends 1000 sync packets per second to the standby via a dedicated 1Gbps link, including real-time process values and logic states. If the primary detects a fault (CPU error, power loss, I/O communication failure), the standby activates in <5ms—faster than the response time of most industrial actuators. Unlike legacy controllers, it uses “stateful failover”: the standby doesn’t just take over, it resumes exactly where the primary left off, avoiding process transients.
Its IEC 61850 stack supports GOOSE messages with 1ms transmission latency—critical for tripping turbine breakers or isolating faulted equipment. The 2.5kV AC isolation between CPU and I/O protects sensitive electronics from high-voltage transients (common in power plants), while the -10°C to +65°C operating range handles unconditioned control rooms in desert or arctic locations. This isn’t just a controller—it’s a promise that critical processes stay online, even when components fail. For plants where downtime costs $10k+ per minute, that’s not a luxury—it’s a necessity.
Redundancy is its defining feature. Two PM864s operate in hot standby: the primary CPU sends real-time status updates to the standby via a dedicated sync link (100Mbps). If the primary detects a fault (e.g., power loss, processor error), the standby takes over in <10ms—faster than most process dynamics, so the loop never deviates. The CPU communicates with I/O modules via a 1Gbps backplane, ensuring fast data exchange with sensors and actuators.
Ruggedization is built in: conformal-coated PCBs resist oil and dust (critical for refineries), and the -20°C to +60°C operating range handles unconditioned control rooms in arctic or desert locations. Its IEC 61850 stack supports GOOSE messages for real-time control (e.g., tripping a breaker in 1ms) and MMS for data logging—integrating seamlessly with smart grid and Industry 4.0 systems. This isn’t a “one-size-fits-all” CPU; it’s engineered for the moments when every millisecond and every redundant component counts.
