Power Management ICs (PMIC) — LDO / DC-DC / Protection / Supervisors

October 09 2025
Ersa

Selection guide for PMICs—LDO vs buck/boost, eFuse/load switch, supervisors/reset, references, sequencing, and USB-PD power path. Small-batch friendly.

What is a PMIC? — System-level power coordinator

A PMIC coordinates multiple rails—conversion, protection, supervision and sequencing—so a board powers up cleanly and safely.

  • Voltage regulation
  • Battery management
  • Sequencing
  • Thermal
  • Power path
Multi-rail PMIC map: input battery/adapter with eFuse/ideal-diode protection feeding a PMIC; three rails buck→CPU, LDO→sensor/ADC, buck-boost→RF/IO; PG/RESET and sequencing arrows
Fig.A — Multi-rail power map (input → protection → PMIC → buck/LDO/buck-boost rails with PG/RESET & sequencing).

What a PMIC does

  • Conversion: buck / boost / buck-boost / LDO to generate stable rails.
  • Protection: eFuse, load-switch, ideal-diode / ORing, surge control.
  • Supervision: voltage supervisors, reset & watchdog for brownout immunity.
  • References: low-noise/low-drift Vref for ADC/DAC & sensors.
  • Sequencing & power path: PG chains, tracking, USB-PD source/sink.

Boundaries & common confusions

  • PMIC vs single regulator: PMIC handles multi-rail coordination & protection.
  • PMIC vs Battery-management IC: BMS manages charge/fuel-gauge; PMIC manages board-level rails.
  • When a PMIC isn’t needed: single rail, low power, no strict sequencing → single LDO/buck may suffice.
Takeaway: PMIC is the board’s power coordinator—it enforces order, safety and predictability across multiple rails, rather than being “just one regulator.”

LDO vs Buck/Boost — When to choose which

Efficiency/thermal and noise/transient pull in different directions. Use this quick math + decision map to make a safe first pick. For quiet rails choose LDO; for heavy loads or big voltage gaps, choose buck/boost; combine buck→LDO for “efficient yet quiet”.

Efficiency/thermal and noise/transient pull in different directions...

LDO thermal loss
Ploss = (VIN − VOUT) × IOUT
Buck effective loss
Ploss ≈ (1 − η) × VIN × IOUT
LDO minimum input
VIN,min = VOUT + VDROPOUT

Rule-of-thumb thresholds

  • If Ploss(LDO) ≳ 0.5–1.0 W or expected efficiency < 70–80% → pick a buck.
  • If IOUT ≲ 50–150 mA and rail is noise-sensitive → pick an LDO.
  • For “efficient yet quiet”, use buck → LDO (front-end buck feeds a low-noise LDO).

Noise & transient checklist

  • Ripple/Noise: LDO is quiet; buck needs proper LC and layout.
  • PSRR: LDO wins at mid-high frequencies; buck injects switching ripple.
  • Load-step: Buck requires compensation & loop bandwidth; LDO depends on COUT & ESR range.
Decision map for LDO vs buck by heat/efficiency, noise and load current; sweet spot shows buck feeding LDO for efficient yet quiet rails
Fig.B — LDO ↔ Buck decision map (heat/efficiency • noise • load current). The “buck→LDO” area yields quiet rails with high efficiency.

Minimal chain — MCU sensor rail (quiet first)

VIN 5V → LDO 3.3V → Sensor/ADC · Pick a low-noise LDO, check PSRR @100k–1 MHz, place decoupling close to loads.

Minimal chain — Wireless rail (efficiency first)

VIN 5V/USB → Buck 3.3V → RF/IO · Control EMI via loop area & grounding; add a 50–100 mA LDO post-filter if the RF is sensitive.

Power Protection ICs — eFuse / Load-Switch / Ideal-Diode / Surge

Ports and inputs fail for the same costly reasons: short, reverse, surge and uncontrolled inrush or backfeed. Protection ICs add controlled limits, timing and directionality that MOSFET-only tricks can’t guarantee.

Quick decision table

Scenario / Need Required capability Recommended IC Placement Key params / notes
On/off with soft-start; quick discharge Controlled slew, output discharge Load-Switch IC Branch / peripheral rail RDS(on), optional ILIM, QOD, slew
OCP/OVP, short-circuit limit, thermal foldback, backfeed block Current/voltage thresholds, fault latch/auto-retry eFuse / Hot-Swap IC Input / port front ILIM, OVP/UVP, TSD, latch, ideal-diode option
Dual-source ORing; reverse-current blocking One-way conduction, fast switchover Ideal-Diode / ORing Ctrl Source merge node Reverse threshold, turn-off time, drop
Automotive/industrial surge & load-dump Over-voltage clamp/foldback, energy handling Surge-Stopper IC Very front of input Transient/steady OVP, current limit, SOA, ISO7637
Inrush quick math
Iinrush ≈ Cload · ΔV / Tramp

Soft-start in load-switch/eFuse increases Tramp. Large capacitive loads without QOD can retain charge and cause overshoot at re-plug. Always size against the source/port limit (USB, adapter, battery).

When QOD is necessary
  • Capacitive loads & hot-plug ports (USB/Type-C).
  • Multiple branches sharing a mid-bus—avoid residual voltage back-bias.
  • Sequenced rails that must return to a known 0 V before restart.
Protection placement & timing: input surge-stopper/eFuse, branch load-switch, USB port eFuse + ideal-diode feeding power-path; faults interact with EN and PG timing
Fig.C — Protection placement and timing across input, branches and USB port; protection acts before PG/RESET.
Three high-stakes rules:
  • USB ports: always eFuse (limit + QOD) and reverse-block (ideal-diode). Verify Iinrush.
  • 12 V automotive: front-end surge-stopper/hot-swap before efficiency tuning.
  • Dual sources: ideal-diode/ORing is mandatory to avoid backfeed and bus dips.

Supervisors & Reset — Brownout Immunity

Random reboots usually stem from rail timing and brownout issues. Pick the right monitor (threshold vs window), set delays/hysteresis, and keep PG/RESET away from switching noise. Add a watchdog where software can stall.

Monitor types

  • Threshold: single limit + fixed delay; simple and cheap.
  • Window: upper & lower bounds; more robust in noisy/multi-rail systems.
  • Watchdog: independent timer to recover stalled software.

Selection checklist

  • Threshold accuracy & temperature drift.
  • Delay range & reset polarity / open-drain.
  • Minimum pulse width and noise filtering.
Start-up and power-down timing windows with PG daisy chain, sequencer controlling EN, and reset release after rails settle
Fig.D — Start-up/shutdown timing windows; PG daisy chain feeds sequencer EN; release reset after rails settle.

Suggested thresholds & delays

Rail Typical threshold Delay suggestion Notes
3.3 V digital 2.9–3.1 V window Power-up 50–200 ms / Power-down 5–20 ms Ignore short dips; prevent nuisance resets.
1.2–1.8 V core 90–95% of Vnom 20–100 ms Match CPU/DDR timing; respect min hold time.
Sensors / analog 95–98% of Vnom 100–300 ms Stabilize analog first; then enable interfaces.

Safe timing templates

  • CPU → DDR → IO: PG daisy chain + fixed delays; shutdown in reverse order.
  • RF → Baseband → Peripherals: bring RF last, drop first to avoid spurs.
  • USB-PD → Core → High-current peripherals: negotiate PD first, then core, then heavy loads.

Validation essentials

  • Capture PG/RESET edges and pulse width.
  • Check threshold stability under load steps.
  • Watchdog timeout policy aligned with software recovery.

Routing rules

  • Keep PG/RESET and EN away from switch nodes and noisy returns.
  • Place dividers near the sensed rail; short, same-layer route.
  • Add RC filtering for remote sense to avoid spike triggers.

Voltage References — Series vs Shunt

For most MCU/SAR-ADC rails, Series references are the safe default. Use Shunt references when wide input range, external clamping or multi-node sharing helps. For ≥20-bit ΔΣ systems, favor low noise/low drift refs with a proper buffer and thermal control.

Series vs Shunt — quick comparison

Dimension Series Reference Shunt Reference
Operation Acts like a regulated source; needs supply headroom. Acts like a precision clamp; sets V by current through a resistor.
Noise (0.1–10 Hz / wideband) Often lower wideband noise; stable with specified C/ESR. Low 0.1–10 Hz parts exist, but layout & bias matter more.
Tempco / long-term drift Low ppm/°C options; good long-term drift families. Widely varies; pay attention to bias current over temp.
Load drive / buffering Limited drive; buffer for heavy/fast-changing loads. Almost always buffer for stability and line isolation.
Input range / special uses Needs VIN > Vref + headroom. Wide range; useful for clamps & shared mid-nodes.
Cost / power Efficient at steady bias; fewer passives. Resistor current wastes power; flexible but cost in bias.
Noise reading & quick math

Check nV/√Hz and 0.1–10 Hz noise. For a rough wideband estimate:

Vn,rms ≈ en × √BW

For 16-bit SAR, BW is typically tens of kHz. Keep Vn,rms ≲ LSB/2 including ADC front-end filtering.

Drift budgeting

Budget ppm/°C × ΔT along with initial accuracy and long-term drift. A thermally tight layout (short Kelvin sense) reduces gradients; re-calibrate for high-accuracy systems.

When to add a buffer

  • Multiple ADC channels or large sampling capacitor.
  • DAC load varies with code or external circuitry.
  • Remote reference routing; need isolation and stability.

Buffer amplifier picks

  • Low noise, low bias, rail-to-rail I/O if needed.
  • Stable with capacitive loads; add small series R + C low-pass.
  • Use a star-point ground; Kelvin sense the reference node.
Series vs Shunt reference concept: series as regulated source with headroom; shunt as precision clamp with bias resistor; buffer and RC filter driving ADC/DAC
Fig.E — Series vs Shunt concepts and a typical buffered Vref feed for ADC/DAC.
Practical thresholds:
  • 16-bit SAR “good enough”: 0.1–10 Hz noise ≤ 1–3 µVpp; integrated wideband noise ≲ LSB/2; tempco ≤ 10–20 ppm/°C.
  • 24-bit ΔΣ: prioritize low 0.1–10 Hz noise and long-term drift; always buffer + RC filter; allow warm-up and periodic calibration.

Power Sequencing & Tracking

Multi-rail boards fail when order, hold time or residual charge is overlooked. Use proven tracking modes, PG daisy chains and soft-start to shape safe ramps; handle pre-bias and dual sources with ideal-diode/ORing.

Tracking modes — when to use which

Mode Benefits Risks Typical use
Coincident Same slope, rails arrive together; simple coordination. Cross-coupling can bite if one rail lags or droops. Core + IO together; some analog domains.
Ratiometric Maintains voltage ratios; gentle for interfaces. Needs careful ramp control and sensing. Mixed-signal ADC/DAC, sensor interfaces.
Offset Enforced order and hold times; robust with PG gating. Longer start-up; tune delays to avoid brown-in. CPU → DDR → IO; RF → Baseband → Peripherals.
Three-rail templates: Offset (CPU→DDR→IO), Coincident and Ratiometric with PG, EN, soft-start, pre-bias handling and reverse-order shutdown
Fig.F — Three-rail templates with PG daisy chain, soft-start, pre-bias and reverse-order shutdown.

PG daisy chain & gating

  • Upstream PG gates downstream EN; add de-glitch & min high-time.
  • Release reset only after all required rails are within window.
  • Mirror PG to a monitor pin for validation captures.

Soft-start, pre-bias & ORing

  • Program Trise to limit Iinrush; coordinate across rails.
  • Ensure buck supports pre-bias start or add ideal-diode to prevent reverse conduction.
  • Use ideal-diode/ORing for dual sources; integrate fault signals into timing.

“Replacement switches” — practical swaps

  • Buck ↔ LDO: keep PG/EN timing & Trise; verify downstream remains unchanged.
  • Single-source → Dual-source: insert ideal-diode/ORing; include its PG/fault into the chain.
  • Load-switch → eFuse: preserve EN logic; re-size ILIM/QOD; re-check inrush.

Bring-up validation

  • Capture PG/EN/RESET ordering at start-up and power-down.
  • Measure Trise, Iinrush and minimum hold times.
  • Brown-in/out and hot-plug behavior under worst loads.

Common pitfalls

  • PG glitches used directly as EN without filtering.
  • Pre-bias ignored, causing reverse conduction at start-up.
  • Dual-source ORing without integrating fault/PG into timing.

USB-PD & Power Path (Type-C)

Ports fail for the same reasons: backfeed, hot-plug surges, wrong roles, or over-current on non-eMarker cables. ≤5V/3A: Type-C is fine with proper port protection (eFuse + reverse block). 20V/5A: USB-PD negotiation + 5A eMarker cable + power-path control are mandatory.

Roles & capabilities — quick reference

Role CC detect (Rp/Rd) BC1.2 USB-PD (PDO/Swap) VCONN / eMarker Power-Path focus Protection focus
Source Advertise via Rp Charge profile @5V Offer PDOs; PR/DR swap May supply VCONN; must gate 5A only with eMarker VBUS sourcing, current limit, thermal eFuse + ideal-diode (reverse block), TVS as needed
Sink Detect via Rd Identify legacy sources @5V Request PDOs; PR/DR swap Check cable current capability VBUS sinking, inrush control, dead-battery eFuse (inrush/short), reverse block to mid-bus
DRP (dual-role) Dynamically swap Rp/Rd Legacy fallback when no PD PR/DR swaps, role policy VCONN control for cable ID Bidirectional power-path with prioritization Symmetric protection in both directions

BC1.2 vs USB-PD — when to use

  • BC1.2: 5V legacy charging profiles (≈0.5–1.5A). No 9/15/20V negotiation.
  • USB-PD: Negotiates PDOs (5/9/15/20V…) and current; enables DRP, role/data swaps.
  • Above 3A@5V or any voltage >5V → use PD and validate cable capability.

Power-Path controller — key duties

  • Path management: source/sink switching, priority, merge.
  • Protection: cooperate with eFuse/ideal-diode (OV/OC, reverse block, thermal foldback).
  • Dead-battery mode; VCONN supply for eMarker; FAULT/PG/INT reporting.
Cable & eMarker: ≤3A: non-eMarker cable allowed.   >3A up to 5A: 5A eMarker cable required. Source must gate 5A only when cable advertises 5A; Sink must still limit current and manage heat.

Port protection essentials

  • eFuse for current limit + inrush (QOD recommended);
  • Ideal-diode or reverse-blocking path to prevent backfeed;
  • ESD on CC/VCONN, selective TVS on VBUS; short, low-impedance returns;
  • Coordinate FAULT with PD/power-path logic to avoid re-strike.

Layout & validation highlights

  • Compute inrush; tune soft-start vs source capability.
  • Measure PDO transitions, PR/DR swaps, and temperature rise at worst case.
  • Check non-eMarker cable behavior (limit current) and reverse insertion.
USB-C port with eFuse current limit and ideal-diode reverse block feeding a PD power-path controller; VCONN to eMarker; protection before PG/EN timing
Fig.G — Type-C/USB-PD power path with eFuse and ideal-diode protection feeding a power-path controller; VCONN/eMarker shown.

Minimal compliant chains — 5V/3A and 20V/5A

Profile Protocol Cable requirement Controller blocks Protection blocks Validation focus
5V / 3A BC1.2 or PD @5V (fixed PDO) Non-eMarker OK Type-C CC detect, optional PD, simple power-path eFuse (inrush/OC) + ideal-diode reverse block; ESD/TVS as needed Iinrush, steady-state temp, unplug/re-plug behavior
20V / 5A USB-PD with negotiated PDOs; PR/DR swap as policy 5A eMarker cable required PD controller + bidirectional power-path; VCONN supply High-current eFuse (QOD) + ideal-diode; robust thermal & surge handling PDO transition transients, 5A thermal budget, role swaps, fault timing

Tip: Validate PD logs (PDOs, PR/DR swaps), then correlate with oscilloscope captures of VBUS ramp, inrush and port thermals at worst-case.

Quick Shortlist — Cross-brand picks

A practical “3 picks per category” to move from selection to quoting. Each table lists a conservative, a performance and a rugged/AEC-Q option. Replace the placeholder PNs with your in-stock parts before publishing.

LDO Regulators

3 picks
Slot PN / Brand VIN / VOUT IOUT Package Highlights AEC-Q
Conservative TI-LDO-X1 / TI 2.7–6 V → 3.3 V 300 mA SOT-223 / SOT-23 Low-Iq, stable ESR window N
Performance ST-LDO-UQ / ST 1.8–5.5 V → 1.8 V 500 mA DFN/QFN Ultra-low noise, high PSRR N
Rugged / AEC-Q REN-LDO-AQ / Renesas 3–40 V → 5 V 150 mA SOIC / HTSSOP Wide VIN, thermal, watchdog option Y

Tip: Analog rails first—low noise & PSRR; ensure dropout headroom.

  • VDROPOUT margin
  • Output cap ESR window
  • RθJA thermal budget

Buck Converters

3 picks
Slot PN / Brand VIN / VOUT IOUT Package Highlights AEC-Q
Conservative NXP-BUCK-S1 / NXP 4–24 V → 5 V 2 A SOIC / TSSOP Wide VIN, easy compensation N
Performance TI-BUCK-UQ / TI 2.9–17 V → 3.3 V 3 A QFN micro High η, spread-spectrum EMI N
Rugged / AEC-Q ON-NCP-AQ / onsemi 6–40 V → 5/3.3 V 2 A SOIC / DFN Automotive transients, hiccup OC Y

Tip: High current or big ΔV → buck; validate compensation & EMI early.

  • Loop compensation
  • Switch loop area
  • Hot-spot thermal

Boost / Buck-Boost

3 picks
Slot PN / Brand VIN / VOUT IOUT Package Highlights AEC-Q
Conservative MCH-BST-S1 / Microchip 1.8–5 V → 5/12 V 0.6 A @5 V MSOP / DFN Simple start-up, internal switch N
Performance TI-BB-UQ / TI 2–18 V → 3.3–12 V 2 A (buck-boost) QFN Seamless mode, high η N
Rugged / AEC-Q ON-BST-AQ / onsemi 3–36 V → 12 V 1 A SOIC / DFN Cold-crank, load dump tolerant Y

Tip: Battery→5/12 V rails; ensure start-up under load.

  • Cold start/UVLO
  • Peak current limits
  • Sync vs diode topology

Protection (eFuse / Load-Switch / Ideal-Diode / Surge)

3 picks
Slot PN / Brand VIN ILIM / IPATH Package Highlights AEC-Q
Conservative MCH-LSW-S1 / Microchip (Load-Switch) 1.8–5.5 V 1–2 A SOT / DFN Soft-start, QOD N
Performance TI-EFUSE-UQ / TI (eFuse) 2.7–18 V 2–3 A QFN OCP/OVP, latch/auto-retry, rev-block N
Rugged / AEC-Q REN-ID-AQ / Renesas (Ideal-Diode/ORing) 4–36 V 3–5 A path SOIC / DFN Fast reverse block, surge-ready Y

Tip: Port/input first—eFuse + reverse block; compute inrush.

  • ILIM & QOD
  • Reverse current block
  • ISO/USB compliance

Voltage Supervisors / Reset

3 picks
Slot PN / Brand Threshold Delay / Features Package Highlights AEC-Q
Conservative MCH-SV-S1 / Microchip 3.0 V (±1.5%) 140 ms; push-pull SOT-23 Simple, robust N
Performance TI-SV-WIN / TI Window (±2%) Adjustable, WDT QFN Noise-immune, watchdog N
Rugged / AEC-Q REN-SV-AQ / Renesas Configurable window Long delay, open-drain TSSOP Brownout immunity Y

Tip: Window + delay to avoid nuisance resets.

  • Threshold accuracy & drift
  • Delay range
  • Reset polarity

Voltage References

3 picks
Slot PN / Brand Vref Noise / Tempco Package Highlights AEC-Q
Conservative ST-REF-S1 / ST (Series) 3.0 V Low / 20 ppm/°C SOT-23 Stable with C, easy drive N
Performance TI-REF-UQ / TI (Series) 2.5 V Ultra-low / 10 ppm/°C QFN Low 0.1–10 Hz noise N
Rugged / AEC-Q REN-REF-AQ / Renesas (Shunt) 2.5 V Low / 15 ppm/°C SOIC Clamp use, wide bias Y

Tip: Series for most SAR; buffer for ΔΣ & multi-load.

  • 0.1–10 Hz noise
  • ppm/°C drift
  • Buffer stability

Sequencers / Trackers

3 picks
Slot PN / Brand Rails / Mode Controls Package Highlights AEC-Q
Conservative TI-SEQ-S1 / TI 2–4 rails / Offset PG chain, fixed delays TSSOP Simple, robust timing N
Performance ST-TRK-UQ / ST 3–6 rails / Coincident, Ratio Programmable ramps QFN Smooth tracking N
Rugged / AEC-Q ON-SEQ-AQ / onsemi 2–4 rails / Offset Fault gating, reset SOIC Automotive timing Y

Tip: Pick template: Offset / Coincident / Ratiometric.

  • PG de-glitch
  • Trise alignment
  • Reverse-order shutdown

USB-PD / Power-Path

3 picks
Slot PN / Brand Role / Voltage Path current Package Highlights AEC-Q
Conservative MCH-USBC-S1 / Microchip Sink 5/9 V (PD) 3 A QFN Dead-battery, simple path N
Performance TI-USBC-DRP / TI DRP 5–20 V (PD) 5 A path QFN PR/DR swaps, VCONN/eMarker N
Rugged / AEC-Q ON-USBC-AQ / onsemi Source 5–20 V (PD) 5 A path QFN / SOIC Reverse block + eFuse coop Y

Tip: 20V/5A needs PD + 5A eMarker + reverse block.

  • PDO logs
  • Iinrush & thermals
  • Cable capability
Specs vary by sub-variant — verify latest datasheet before build. Submit your BOM — 48h, 3 safe cross-brand options

Troubleshooting — Problem → Cause → Verify → One-shot Fix

Match the symptom → capture proof under “Verify” → apply the “One-shot Fix”. Always verify first before changing hardware/firmware. Related topics to link later: Protection / Supervisor / Buck / Sequencer.

Power-up oscillation / repeated start attempts

Common
Problem Likely causes Verify (scope) One-shot fix
VOUT rises then collapses in cycles Soft-start too fast → Iinrush exceeds source; unstable compensation; output cap ESR outside stable region; pre-bias not handled CH1=VOUT, CH2=IOUT, CH3=EN/PG; trigger on VOUT falling edge (repeating); watch current-limit/thermal foldback Lengthen soft-start / lower current limit; correct compensation & ESR; enable pre-bias start or add ideal-diode ahead

→ Link later: Buck / Protection / Sequencer

Brownout (voltage dip causes reset)

Supervision
Problem Likely causes Verify (scope) One-shot fix
Reset after load step / supply dip Supervisor thresholds/delays too tight; PG directly drives EN without de-glitch; main rail sags (trace resistance/ground return) CH1=VCORE, CH2=3V3, CH3=RESET, CH4=PG; trigger on RESET falling edge; measure dip depth & duration Use windowed supervisor and relax delay; PG → Sequencer → EN daisy chain; add local decoupling & improve ground return

→ Link later: Supervisor / Sequencer

Overshoot (power-up / unload exceeds spec)

Loop/Layout
Problem Likely causes Verify (scope) One-shot fix
VOUT exceeds upper tolerance Excess loop bandwidth/underdamped; rapid unload; synchronous rectifier back-feed; noisy FB/Kelvin routing Trigger when VOUT > limit; log amplitude & duration; compare with BW limit ON/OFF Retune compensation; add slight ESR/RC; small RC snubber at load; clean FB/Kelvin routing

→ Link later: Buck

Port current-limit false trip (drops right on plug-in)

Type-C/PD
Problem Likely causes Verify (scope) One-shot fix
Power cuts immediately at plug-in Iinrush not limited; too-narrow detection window; 5A allowed without 5A eMarker; reverse current transient flagged as fault CH1=VBUS, CH2=IOUT/Rsense, CH3=FAULT, CH4=CC/VCONN; trigger on FAULT falling edge; check slope & peak eFuse soft-start + QOD; de-glitch current limit; allow 5A only with 5A eMarker cable; ideal-diode for reverse-block

→ Link later: Protection / USB-PD & Power-Path

Load-Dump / automotive surge damage

Automotive
Problem Likely causes Verify (scope) One-shot fix
Parts fail or power loss after surge No surge-stopper/hot-swap; TVS energy too low; SOA exceeded Surge generator or equivalent; CH1=VIN, CH2=VOUT, CH3=IIN, CH4=FAULT; trigger when VIN > threshold; log SOA & thermals Add surge-stopper (clamp/foldback); higher-energy TVS; derate & recheck thermal and return paths

→ Link later: Protection

Oscilloscope trigger presets

  • Power-up oscillation: Edge (fall) @ VOUT = 0.9·Vnom, holdoff 2–5 ms, BW limit ON, ≥100 MS/s.
  • Brownout: Pulse-width (low) on VCORE @ 0.9·Vnom, width > 50 µs; qualify with RESET.
  • Overshoot: Runt (positive) on VOUT, upper=1.05·Vnom, lower=Vnom, time > 10 µs.
  • Port limit false trip: Edge (fall) on FAULT + math/DB for IOUT peak; add CC/VCONN digital trigger.
  • Load-Dump: Edge (rise) on VIN > 1.3×rated; 1–10 ms/div; optional thermal camera sync.

PG/RESET verification — minimal closed loop

  1. Wiring: CH1=PG, CH2=EN, CH3=RESET, CH4=critical VOUT; common ground point.
  2. Startup capture: Record full power-up; check PG → EN → RESET order & pulse widths.
  3. Window check: Confirm supervisor threshold (%Vnom) and delay (ms) are within recommended range.
  4. Power-down: Cut upstream source; verify reverse-order shutdown & minimum hold time.
  5. Disturbances: Apply load steps / hot-plug; watch for false trips.
  6. Conclusion: Log “evidence (waveforms) → chosen fix”, then rerun to confirm.

FAQ — Engineering & Selection

What does a power management IC do?

At board level, a PMIC coordinates multiple rails: regulation (buck/boost/LDO), protection (eFuse/load-switch/ideal-diode), supervision/reset, and sequencing. It turns a battery or adapter into stable rails with predictable startup/shutdown, while controlling inrush, limiting heat, and preventing brownouts that crash CPUs or radios.

→ See: What is a PMIC

What is a power management IC?

A power management IC (PMIC) integrates several power functions—converters, references, protection, and power-path control—into one device or tightly coupled set. Compared with many discrete parts, it reduces area, routing complexity, and timing risk, especially on dense multi-rail systems where sequencing and brownout immunity matter.

→ See: What is a PMIC

What is the IC in a power supply?

In modern designs, “power-supply IC” usually means a switching regulator (buck/boost) or an LDO that sets a regulated rail. Around it, PMICs add supervisors, references, and protection ICs to meet timing, noise, and safety targets—ensuring rails start cleanly and survive faults or hot-plug events.

→ See: LDO vs Buck/Boost

What type of IC is used in voltage regulation?

For higher currents or large VIN–VOUT gaps, buck/boost converters deliver efficiency. For low current or ultra-low-noise rails, LDO regulators shine. Many PMICs cascade both: a high-efficiency buck feeding quiet LDO rails for ADCs, clocks, or radios, balancing ripple, PSRR, thermal budget, and BOM.

→ See: LDO vs Buck/Boost

How to test a power IC?

Probe VIN, VOUT, and PG/RESET. Verify startup ramp, load-step recovery, ripple, and dropout. For protection parts, measure current-limit, short-circuit response, and inrush using C·ΔV/Tr. Use a short coax ground, enable bandwidth limit, and capture brownout or fault-latch behavior for repeatability and evidence.

→ See: Troubleshooting , Power protection

How to know if power IC is damaged?

Red flags include abnormal quiescent current, VOUT stuck low/high, persistent thermal shutdown, or PG never asserting. Check for output shorts and isolate loads. If the IC overheats unloaded or fails soft-start repeatedly, replace it, then review ratings, compensation, layout, and heat spreading around the package.

→ See: Troubleshooting

Can power IC be replaced?

Yes—start with pin-compatible families. If not, match control behavior (EN/PG logic), compensation, soft-start profile, OCP/OVP thresholds, and thermal pad footprint. Re-validate sequencing with other rails and confirm brownout immunity and inrush limits before committing the alternate to EVT/DVT builds or field repairs.

→ See: Quick shortlist (CTA: Submit your BOM)

How to check if the power supply is faulty?

Load the input and watch droop, verify loop stability (no oscillation on steps), and check sequence order relative to CPU/DDR rails. Brownouts or bouncing during inrush indicate soft-start, eFuse/load-switch, or PG thresholds need tuning. Capture waveforms and correlate with FAULT/RESET pins for root cause.

→ See: Troubleshooting , Power sequencing/tracking , Power protection

What is IC regulated power supply?

It’s a power stage built around integrated regulators—buck/boost or LDO—to keep a rail within tolerance across load and VIN variation. Compared with linear transformer bricks, IC-based supplies enable compact multi-rail boards with sequencing, supervision, and protection to meet thermal and reliability goals.

→ See: What is a PMIC , LDO vs Buck/Boost

What is IC voltage?

In PMIC context, it means allowable VIN/VOUT ranges and internal reference levels. Check absolute maximums, dropout and headroom for LDOs, and duty-cycle/ton limits for converters. Ensure margins across tolerance and temperature, and verify PG thresholds align with your rails’ safe operating windows.

→ See: LDO vs Buck/Boost

What does power management do?

Power management balances efficiency, noise, and timing. It sequences rails, protects ports and batteries, enforces current/voltage limits, and keeps thermals within budget. A well-planned PMIC architecture reduces field failures, simplifies second-sourcing, and provides predictable startup/shutdown even under hot-plug or brownout stress.

→ See: What is a PMIC , Sequencing

How to use IC in a circuit (for power rails)?

Start from loads: voltage, current, noise, and transients. Choose buck/boost/LDO accordingly, then add input protection (eFuse/ideal-diode) and supervisors for critical rails. Keep switch loops tight, place input/output capacitors close, separate PG/EN from noisy nodes, and validate sequencing templates against CPU/DDR needs.

→ See: LDO vs Buck/Boost , Power protection , Sequencers/trackers

Which IC is the best?

There’s no universal “best.” Quiet analog rails favor low-noise LDOs; heavy loads demand buck/boost; harsh inputs need eFuses or surge-stoppers. Build a cross-brand shortlist (3 parts), then verify timing, inrush, thermals, and protection behavior on your board before freezing the choice for production.

→ See: Quick shortlist

What devices use IC? (PMIC examples)

Phones, wearables, and IoT nodes rely on PMICs for size, battery life, and multi-rail coordination. Automotive ECUs add surge and reverse protection. USB-C hosts/sinks use PD and power-path controllers with eFuses and ideal-diodes to negotiate PDOs cleanly and pass compliance and thermal checks.

→ See: USB-PD & power-path , Power protection ICs

Ready to move from shortlist to action?

Submit your BOM (48h)

Within 48 hours you’ll receive three safe options with indicative lead time, price ranges, and short notes on thermal, inrush, and sequencing risks.

  • Small-batch friendly
  • Pin-compatible first
  • AEC-Q optional
Submit your BOM — 48h • 3 safe options Your BOM is used only for technical evaluation. No marketing blasts. No resale of data.

Not ready to submit?

Start with a concise IC selection guide to frame loads, rails, and protection before you collect quotes.

Read the IC selection guide
Ersa

Anastasia is a dedicated writer who finds immense joy in crafting technical articles that aim to disseminate knowledge about integrated circuits (ICs). Her passion lies in unraveling intricate concepts and presenting them in a simplified manner, making them easily understandable for a diverse range of readers.