OR-Ing Ctl
Continuously compares input and output voltages of an N-FET acting as an OR-ing diode, driving the gate fully on for forward current and slamming it off in <1 µs when reverse current would flow because of an input failure.
A high-speed comparator watches VDS across the FET; when VDS goes negative (forward current), the controller's charge pump turns the gate fully on, dropping VDS to RDSon × ILoad. When VDS approaches zero or reverses, the chip detects pending reverse current and pulls the gate to source within ~1 µs, isolating the failing input. Some parts add fault flags and 'CR' (current-regulating) modes for inrush control.
In plain terms
Like a one-way valve in a city's water main — water from any of three reservoirs flows toward the customer, but if one reservoir loses pressure, the valve seals so the others don't drain back into the empty tank.
Why designers use it
- Server-grade redundancy — telecom and data-centre PSUs combine N+1 supplies so one failure doesn't crash the rack.
- Low forward drop — replacing a 0.4 V Schottky with a 5 mΩ MOSFET saves 4 W per 10 A, which is the difference between a fan-cooled and convection-cooled 80+ Titanium PSU.
- Fast reverse blocking — under 1 µs response prevents an upstream input cap from momentarily reverse-charging through the failed supply.
- Ground-referenced control — the chip sits on the load side and needs no isolated bias.
Best for
- Redundant PSUs
- Battery-AC backup
- USB-C source select
Key specifications
- Operating limits: Check the exact manufacturer's datasheet (A family name does not establish voltage, current, temperature or timing limits.)
- Pin assignment: Match the complete part and package code (A similar name or function does not guarantee the same wiring.)
- Mechanical fit: Use the exact package drawing (Check pad layout, dimensions and viewing direction before building.)
When not to use it
- For low-current (<2 A) systems — a Schottky's 0.4 V loss is acceptable and a controller plus FET costs more than the diode it replaces.
- When AC input combining is needed — OR-ing controllers are DC parts; AC needs a different topology.
Common mistakes
- Choosing an FET with too high RDSon — the controller forward-biases the FET correctly, but the I²R loss climbs and the FET runs hot.
- Placing the input and output sense Kelvin pins on top of the high-current trace — induced voltage shifts trip the reverse-current comparator on every step-load change.
Where you will find it
- A Cisco 4500X switch's redundant PSU shelf uses an Analog Devices LTC4351 OR-ing controller per supply driving an Infineon BSC079N03 FET: a yanked PSU is isolated within 700 ns, before the input cap can reverse-charge from the surviving rails, keeping the switch fabric alive without a packet drop.
- A Tesla Powerwall battery backup uses a Texas Instruments LM5050 OR-ing controller to combine grid power and battery output onto the inverter bus: when the grid drops, the controller turns its FET off in 500 ns and the battery's matching FET picks up, providing a glitchless transition that keeps a connected fridge from cycling.
- A USB-C laptop charger negotiates dual upstream PD sources through a TI LM5050-1: the chip seamlessly hands off between an 87 W and a 65 W supply when the user swaps the brick mid-flight, so the laptop never falls back to battery during the 200 ms reconnect window.
A short history
OR-ing controllers manage the transition between two or more redundant power supplies in N+1 or fully redundant systems, ensuring the load current transfers seamlessly from a failing supply to a healthy one without any voltage glitch. OR-ing was traditionally performed with Schottky diodes, but their ~0.4 V forward drop wastes power. Linear Technology (LTC4351) and Texas Instruments introduced MOSFET-based OR-ing controller ICs around 2000, replacing diodes with active MOSFET switches that drop millivolts at full load. These ICs are standard in server and telecom power systems where downtime is not acceptable and power efficiency is paramount.