Hot-Swap Ctl
A hot-swap controller controls a series MOSFET to ramp the supply voltage onto a newly-inserted board, limiting the inrush current that the board's bulk capacitors would otherwise demand. Once the board is fully powered and current is within limits, it asserts a power-good signal to begin operation.
A hot-swap controller manages a series MOSFET so connecting an uncharged input capacitor does not demand uncontrolled inrush current. One approach controls voltage rise, using I = C × dV/dt; another measures current through a sense resistor and regulates it directly. Fault thresholds, timing and retry behavior depend on the selected controller. Check the MOSFET's safe operating area throughout startup and faults, not only its steady-state power rating.
In plain terms
An airlock for electricity: it slowly pressurises the new module to match the backplane voltage before fully connecting it, so neither side sees a sudden shock.
Why designers use it
- Allow telecom line cards to be inserted without disrupting 48 V bus voltage for other cards.
- Enable live insertion of storage drives and PCIe cards in enterprise servers.
- Protect backplane traces from the surge that bypasses polyfuses and confuses e-fuses.
Best for
- Telecom / ATCA
- Server blades
- 48 V bus
Key specifications
- Operating voltage: 2.7 V – 80 V
- Programmable current: 0.1 A – 100 A
- Inrush dV/dt control: 10 V/ms – 1 V/µs
- Response time: 1 µs – 100 µs
- Auto-retry: Latch-off / programmable
When not to use it
- When the selected controller's supply, current-sense or gate-drive range does not fit the circuit.
- When the chosen MOSFET cannot remain within its safe operating area during startup or a fault.
- When a simpler load switch already meets the verified inrush and protection requirements; this is an application choice, not a universal 5 V cutoff.
Common mistakes
- Under-sizing the sense resistor and missing the trip point due to its self-heating shifting resistance.
- Forgetting to place the MOSFET on the high-side — a low-side MOSFET cannot block the board's own capacitor pre-charged to rail voltage.
Where you will find it
- A data-centre switch allows a line card to be swapped without power-cycling the chassis: the hot-swap controller on the card's input ramps the 48 V rail onto the card's 10 000 µF bulk capacitors over 5 ms, keeping the inrush below the 20 A per-slot limit specified in PICMG 3.0.
- A submarine sonar rack uses hot-swap controllers on every sub-module so a failed DSP board can be replaced at sea: the controlled ramp prevents the surge from inducing noise in the sensitive hydrophone amplifiers on adjacent slots.
- A CERN particle detector's front-end electronics board uses a hot-swap IC to power-sequence its FPGA, ADC, and analog rails: the built-in power-good sequencing ensures the FPGA's core rail rises before its I/O rail, preventing latch-up in the gate oxide during startup.
A short history
A hot-swap controller manages the inrush current when a board connects to a live supply. Many designs control an external MOSFET and monitor current as the board's input capacitors charge. They can also disconnect an excessive load. Correct component selection and timing remain important; the controller alone does not guarantee a harmless connection.
Good to know
- An uncharged capacitor can briefly look like a heavy load when a card first meets a powered backplane.
- Controlling voltage rise and directly regulating current are different ways to manage inrush.
- A short power pulse is not automatically safe: voltage, current, pulse duration, temperature and the MOSFET's documented safe operating area must be considered together.