123 Main Street, New York, NY 10001

Submit Your BOM

RTC Backup & Switchover (Coin-Cell & Supercap Holdover)

RTC Backup & Switchover (Coin-Cell & Supercap Holdover)

← Back to:Reference Oscillators & Timing

RTC Backup & Switchover is about keeping time and critical retention alive through power loss with ultra-low leakage and a glitch-free, no-backfeed supply transition. This page provides a practical template—leakage budget → energy choice → switchover topology → validation/production criteria—so holdover and integrity targets are met in the real world.

What “RTC Backup & Switchover” Really Means (Scope + Success Criteria)

RTC backup is not only a battery pin. It is a backup domain that must preserve time-related state under a secondary source, while switching between rails without reset, corruption, or untraceable events. This page focuses on energy source selection, ultra-low leakage, and switchover integrity—with measurable pass criteria.

In-scope (this page)
  • Backup domain power path: VBAT pin, switchover block, ORing/ideal-diode/PFET paths.
  • Backup source: coin-cell / supercap / hybrid holdover and protection (within backup boundary).
  • Ultra-low leakage budgeting and isolation: device + board + protection + measurement paths.
  • Switchover integrity verification: waveforms, flags, and event traceability.
Out-of-scope (linked, not expanded here)
RTC calendar/register basics Secure RTC / signed time / tamper TCXO-disciplined RTC accuracy

Those topics can be referenced elsewhere to prevent sibling-page overlap and repeated explanations.

The three KPIs that define success

Every design choice (source, topology, layout, test) must close these KPIs with explicit conditions and pass criteria.

Backup current
  • Definition: average VBAT draw under backup mode.
  • Condition: include temperature points (room + hot).
  • Pass: below target budget, no order-of-magnitude jump with temperature.
Holdover time
  • Definition: duration that the “must survive” list remains valid.
  • Model: coin-cell (usable capacity) or supercap (C·ΔV/I).
  • Pass: meets target with end-of-life and temperature guardbands.
Switchover integrity
  • Definition: VCC↔VBAT switching without reset/corruption.
  • Check: capture VBAT waveform and a switchover/power-fail flag.
  • Pass: VBAT stays above minimum retention level; events are traceable.
Minimum deliverable outcome (what this page enables)
  • A leakage budget worksheet that identifies each contributor and a method to isolate it.
  • A topology decision map (coin-cell / supercap / hybrid + ORing approach) tied to the three KPIs.
  • A verification plan that prevents false readings (meter burden, fixture leakage, humidity effects).
Common failure modes this page targets
  • Holdover is much shorter than calculations due to hidden leakage paths (board, protection, measurement).
  • Switchover “works” electrically but triggers RTC reset because VBAT dips during fast VCC transients.
  • Production screening cannot reproduce lab results because fixtures and meters dominate nA–µA behavior.
Figure 1 — Page boundary map (backup domain, rails, and KPIs)
RTC Backup & Switchover — Page Boundary Map Diagram shows primary rail VCC and backup source feeding an RTC backup domain through a switchover path, with three KPI boxes: backup current, holdover time, and switchover integrity. Primary Rail VCC Backup Source Coin-cell / Supercap RTC Backup Domain RTC core Retain RAM TS ctr Switchover VBAT KPI: Ibackup nA–µA KPI: Holdover h–months KPI: Switch OK no reset Out-of-scope: RTC basics Out-of-scope: Secure RTC Out-of-scope: TCXO-RTC

Requirements Decomposition: Holdover, Events, and “What Must Survive”

Requirements define the backup domain. The key is not “power remains,” but that specific state stays valid and events remain traceable across switchover, temperature, aging, and field maintenance. The following decomposition prevents over-design and keeps leakage under control.

Must survive (minimum functional correctness)
  • Time counter: continuous time base under backup (no discontinuity).
  • Calendar/date (if required): human-readable time remains consistent after restore.
  • Power-fail indication: a reliable flag/event that backup mode occurred.
  • Critical retention bytes: only essential state (e.g., calibration-valid, last-good timestamp).
Failure cost: system time becomes invalid, event logs become non-correlatable, field diagnosis becomes ambiguous.
Nice to have (diagnostics and maintenance efficiency)
  • Switchover count: identifies frequent rail instability and predicts battery wear.
  • Battery-low history: duration below threshold helps distinguish aging vs transient dips.
  • Alarm queue under backup: only if wake-on-alarm is required by the product.
Failure cost: product still works, but root-cause and service workflows become slow and expensive.
Not required (avoid leakage inflation)
  • Large logs or high-frequency telemetry that forces continuous measurement under VBAT.
  • Always-on dividers, LEDs, or protection parts with high reverse leakage at temperature.
  • Non-critical state that can be recomputed or restored after power returns.
Failure cost: none—these should be excluded to protect the backup current budget.
Event requirements (traceability without adding leakage)
  • Power-fail timestamp: enables log alignment and offline-duration reconstruction.
  • Brownout classification: distinguishes slow decay vs abrupt cut (helps switchover debugging).
  • Battery-low: triggers service warning and can throttle monitoring duty-cycle.
  • Switchover count: highlights unstable rails and prevents “mysterious time drift” reports.

Any monitoring path attached to VBAT must be treated as a leakage contributor. If sensing is required, it should be duty-cycled or gated.

Temperature and storage constraints (design guardbands)
  • Low temperature: coin-cell impedance rises, making VBAT dips more likely during transitions.
  • High temperature: supercap self-leakage and reverse leakage of protection parts can dominate holdover.
  • Storage aging: end-of-life behavior is not linear; guardband must cover usable capacity and leakage drift.
Field maintenance constraints (battery replacement without time loss)
  • Service model: user-replaceable vs sealed changes the acceptable “time discontinuity” policy.
  • Swap method: hot-swap or full-off defines whether a temporary hold path is needed.
  • Risk control: define and validate a safe replacement flow that does not add always-on leakage.
Figure 2 — Survival checklist ladder (must survive vs optional vs excluded)
RTC Backup — Survival Checklist Ladder Ladder diagram with three tiers: Must survive, Nice to have, Not required. Each tier contains small blocks for TIME, RAM, ALARM, LOG, and event traceability tags. Must survive TIME RAM LOG POWER-FAIL correctness Nice to have ALARM SWITCH CNT BAT-LOW HISTORY service Not required Always-on ADC Big telemetry Leaky protect Guardband for temperature, aging, and maintenance flow

System Architecture Options (Where to Place the Switchover)

Switchover placement sets the failure boundary. The choice is not only about schematic simplicity, but also about drop headroom, reverse leakage risk, and how easily issues can be isolated during validation and production.

Decision drivers (keep these fixed across all options)
Drop headroom Reverse leakage Complexity Observability

Each option below is evaluated using the same four drivers to prevent “BOM-first” decisions that later fail holdover or integrity.

Option A — Built-in switchover (RTC/PMIC internal)
  • Pros: fewest parts, shortest path, predictable thresholds and sequencing.
  • Risks: VBAT limits (range, allowable charge/backfeed) can silently reduce usable holdover.
  • Typical failures: (1) VCC backfeeds VBAT and drains coin-cells; (2) fast VCC drop causes VBAT dip and reset.
  • Best fit: moderate holdover targets, low BOM, and acceptable fixed behavior.
Option B — External diode-OR
  • Pros: simplest external implementation; easiest to reason about current flow.
  • Risks: forward drop reduces VBAT headroom; reverse leakage can dominate at temperature.
  • Typical failures: (1) high-temperature leakage shortens holdover; (2) drop pushes VBAT near thresholds and increases reset susceptibility.
  • Best fit: prototypes, narrow temperature range, or non-aggressive holdover goals.
Option C — External ideal-diode / PFET ORing
  • Pros: low drop improves usable window; better control of backfeed and transitions.
  • Risks: reverse leakage and gate biasing must be validated across temperature and corners.
  • Typical failures: (1) PFET body diode orientation causes backfeed; (2) controller IQ/leakage breaks nA–µA budgets.
  • Best fit: tight voltage headroom, long holdover targets, and strict reset intolerance.
Option D — Always-on island (dedicated backup power domain)
  • Pros: strongest isolation and system-level control; supports monitoring and redundancy cleanly.
  • Risks: extra always-on loads easily inflate leakage by an order of magnitude.
  • Typical failures: (1) “AON creep” adds dividers/TVS/LEDs and kills holdover; (2) poor domain partitioning introduces hidden leakage paths.
  • Best fit: carrier/industrial systems requiring observability and robust field workflows.
Quick selection mapping (rule-of-thumb)
  • Tight voltage headroom + long holdover: prefer Option C or D to protect usable VBAT window.
  • Low BOM + moderate targets: Option A is usually the fastest path if VBAT specs and backfeed limits are clean.
  • Prototype + narrow temperature: Option B can be acceptable when drop/leakage are explicitly budgeted.
  • High observability / redundancy needs: Option D supports system-level hooks, but leakage discipline is mandatory.
Figure 3 — Four switchover placement topologies (drop vs leakage)
Switchover Placement — 4-way Topology Panel Four panels show different placements of switchover: built-in, diode OR, ideal diode PFET OR, and always-on island, each annotated with drop and leakage tags. A) Built-in VCC RTC/PMIC VBAT Drop: low Leak: spec B) Diode-OR VCC VBAT D1 D2 RTC VBAT in Drop: higher Leak: temp C) Ideal / PFET OR VCC VBAT PFET PFET Ideal CTRL Drop: very low Leak: validate D) Always-on island AON Rail RTC Keep AON minimal Drop: lowest Leak: discipline

Backup Energy Source: Coin-Cell vs Supercap (Deep Trade-offs)

Holdover is set by usable energy and total leakage, not by capacity nameplate alone. Coin-cells typically win on leakage and long-term storage behavior, while supercaps win on pulse power and rechargeability. Hybrid designs split responsibilities: supercaps handle transients; coin-cells supply long-duration backup.

Coin-cell (low leakage, long holdover)
  • Usable capacity depends on temperature: low-temperature impedance reduces usable VBAT window.
  • Pulse capability is limited: fast transitions can create VBAT dips that trigger resets.
  • Charging is typically prohibited: backfeed and trickle paths must be prevented.
Supercap (pulse power, rechargeable)
  • Self-leakage can dominate: ILKG rises strongly with temperature and may exceed RTC draw.
  • Charge management is mandatory: limit inrush, prevent over-voltage, and isolate reverse paths.
  • Total leakage is system-level: include charger, clamps, and board contamination in I_total.
Hybrid (split transient vs long-duration)
  • Supercap role: absorbs fast transitions and short interruptions without VBAT droop.
  • Coin-cell role: supplies long-term backup with minimal leakage.
  • Key constraint: ensure supercap leakage and charge path do not erase the coin-cell advantage.
Holdover templates (use usable energy and total leakage)
Supercap
Holdover ≈ (Cusable × ΔV) / Itotal
Itotal = Ibackup + Icap_leak + Icharger + Iprotect + Iboard
Coin-cell
Holdover ≈ Capacityusable / Iavg
Capacityusable is reduced by temperature, end-of-life impedance, and minimum VBAT limits.
Figure 4 — Energy vs leakage trade (coin-cell, supercap, and hybrid)
Coin-cell vs Supercap — Energy and Leakage Trade Left block represents coin-cell with low leakage and long holdover; right block represents supercap with high pulse power and higher leakage; middle indicates hybrid splitting responsibilities. Coin-cell BAT Low leakage Long holdover Low pulse Supercap CAP High pulse Recharge Leak vs temp Hybrid BAT CAP pulse long Holdover is set by usable energy AND total leakage

Switchover Topologies in Detail (Drop, Reverse Leakage, and Glitch Risk)

Switchover integrity is determined by three hazards that must be verified at the VBAT pin: drop headroom, reverse leakage, and glitch during transition. The goal is to keep VBAT above the retention threshold and prevent any sustained backfeed that silently drains the backup source.

Switchover hazards (engineering checklist)
Drop Reverse leakage Glitch risk Pin-level verify

A topology is acceptable only if VBAT stays within the valid backup operating window across temperature and repeated transitions.

Diode-OR — Drop (VF) and reverse leakage (IR)
  • Drop headroom: VF reduces VBAT margin and can push VBAT near the retention threshold.
  • Reverse leakage: Schottky IR can rise sharply with temperature and dominate the budget.
  • Typical failure: “holdover looks fine on paper” but collapses at hot due to IR + board leakage.
PFET-OR — Body diode path and gate biasing
  • Body diode direction: an incorrect orientation creates a hard backfeed path (VCC → VBAT).
  • Gate bias: a “half-on” region during transition can cause VBAT dip or unwanted conduction.
  • Typical failure: coin-cell drains rapidly while VCC is present due to unnoticed backfeed.
Ideal-diode controller — Low drop, but watch IQ and drive paths
  • Advantage: low effective drop improves usable VBAT window and reduces transition stress.
  • Hidden cost: controller IQ, sensing network leakage, and gate drive can break nA–µA budgets.
  • Typical failure: drop is solved, yet holdover still misses target due to IQ/leakage at hot.
Why RTC resets during switchover (common root causes)
  • VBAT sag: source ESR/ESL + path resistance + transition current causes a brief dip below retention.
  • Ground bounce: return path disturbance makes the RTC “see” a lower effective VBAT.
  • Protection back-injection: clamps can conduct during transients, creating unintended current paths.
  • Long leads: battery holder and wiring inductance turns fast edges into spikes and ringing.
Minimal verification points (no register deep-dive)
Scope probes
CH1: VCC (or primary rail)
CH2: VBAT at the RTC pin
CH3: ORing node / switchover indicator
Pass criteria
VBAT never crosses the retention minimum
No sustained backfeed into VBAT
Power-fail event and reset flag are consistent
Repeatability
Repeat hot/cold transitions
Include long-lead worst case
Log events with timestamps
Figure 5 — Switchover current flow map (forward, reverse, body diode paths)
Switchover Current Flow Map Block diagram showing VCC source, backup source, ORing block, and RTC backup domain with arrows indicating forward path, reverse leakage risk, and body diode path. VCC Backup Coin / Cap ORing / Switch D PFET IDEAL Sense + control RTC Backup Domain VBAT pin RTC core RAM TS Reverse risk Body diode Probe points CH1: VCC CH2: VBAT pin CH3: ORing node Legend Forward path Reverse leakage risk Body diode path

Ultra-Low Leakage Budget (nA–µA): Where the Current Really Goes

Backup holdover is a system-level leakage problem. The only reliable approach is a closed loop: define a budget, measure the total at the backup source, and isolate contributors until the delta is explained.

Budget definition (source-current view)
Itotal = IRTC(VBAT) + Iswitch + Iprotection + Iboard_leak + Ipull + Imeasurement
Always measure at the backup source and reconcile every term. “RTC current” alone is not a holdover prediction.
Top sneaky leaks (high-frequency offenders)
Flux / contamination
Humidity-sensitive leakage that varies by handling and cleaning quality.
TVS / ESD parts
Reverse leakage can rise at hot and become the dominant contributor.
Dividers / always-on nets
“Small” resistive loads become fatal at nA–µA budgets if left powered by VBAT.
Testpoints / fixtures
Dirty pads, pogo pins, or long leads introduce parallel leakage paths.
Cables / connectors
Surface leakage and insulation resistance issues often appear only after soak time.
Budget template (fill-in, isolate, close the delta)
Fields
Target (budget)
Allocated by term
Measured total
Measured per term
Delta and suspects
Next isolation step
Isolation order
1) Remove/disable protection leaks
2) Disconnect dividers and always-on nets
3) Eliminate fixture/testpoint leakage
4) Clean/soak and re-measure
5) Split ORing/controller terms
Temperature discipline
Measure at room and hot first
Repeat after soak time
Log delta by term
Treat hot leakage as worst-case
Figure 6 — Leakage budget bar + top sneaky leaks
Leakage Budget — Stacked Bar and Sneaky Leaks Stacked bar represents leakage contributors compared to a target line; five icons show common sneaky leak sources like flux, ESD, divider, testpoint, and cable. Leakage budget (stacked contributors) RTC Switch Protect Board Pull Meas Target Measured Close the delta: measure total at source, isolate terms, re-measure at hot. If “Board” grows with humidity/soak, prioritize cleaning and fixture leakage. Top 5 leaks Flux ESD Divider TP Cable / fixture Leak path Soak-sensitive

Charging & Protection (Supercap charge, Coin-cell safety, and Backfeed clamps)

Charging and protection must enforce two constraints at the same time: safety guardrails (no forbidden charge paths, controlled voltage/current) and leakage discipline (the protection network must not break the nA–µA budget).

Direction rules (design must enforce)
Supercap: allowed (with guardrails)
VCC → CAP charging is permitted, but must be current-limited and over-voltage protected. CAP → VBAT discharge is permitted, but reverse isolation must prevent backfeed into VCC.
Coin-cell: forbidden charging path
Any sustained VCC → coin-cell current path is a design hazard. The schematic must block this by construction. Coin-cell → VBAT discharge is allowed for holdover.
Supercap charge guardrails (four must-haves)
1) Current limit
Prevent inrush that droops VCC and reduces switchover integrity. Limit charge current at power-up and after deep discharge.
2) Over-voltage clamp
Ensure VCAP never exceeds the safe ceiling. Treat clamp leakage at hot as a budgeted contributor, not an afterthought.
3) Charge-full detect
Provide a practical observability hook: “cap never reaches full” indicates excessive leakage, incorrect clamp, or charge-path issues.
4) Reverse isolation
Block reverse current paths when VCC falls. Prevent CAP/VBAT from backfeeding VCC or creating a hidden discharge loop.
Protection selection constraints (must not break leakage budget)
  • Any clamp/TVS in the VBAT neighborhood must be evaluated for hot reverse leakage. If leakage can dominate, move protection to a different domain or change the protection strategy.
  • Charge-path components (OV clamp, detect network, controller) must declare an explicit contribution to Itotal in the leakage budget.
  • Series resistance can protect and limit inrush, but must be validated against VBAT dip during switchover and peak load events.
Minimal verification points (charge + leakage + backfeed)
Charging profile
Measure VCAP rise and charge current peak. Confirm VCC does not droop into brownout during worst-case cap recharge.
Leakage at steady state
Measure source current with VCC present and absent. Re-check at hot; clamp/controller leakage often rises with temperature.
Backfeed check
During VCC removal and re-apply, verify no sustained current flows into coin-cell and no CAP/VBAT backfeeds VCC.
Figure 7 — Charge path with guardrails (limit, clamp, isolate, no-coin-cell-charge)
Charge Path with Guardrails Block diagram with VCC charging a supercap through a current limiter, an over-voltage clamp, and an isolation block feeding ORing into VBAT, with a coin-cell branch marked no charge. VCC Rlim Current limit Supercap VCAP node OV clamp VCAP ceiling Isolation No backfeed ORing to VBAT VBAT RTC domain Coin-cell Holdover NO CHARGE Backfeed clamp Legend Forward allowed Reverse risk

Monitoring, Flags, and Event Logging Hooks (Design for Observability)

Observability must be designed into the backup system so field failures can be diagnosed quickly and production screening is repeatable. Monitoring must remain low-power; measurement paths that stay “always-on” can silently become the dominant leakage term.

What to observe (backup-domain relevant signals)
VBAT voltage
Trend and minima indicate battery aging, unexpected leakage, and dip events during switchover.
Battery-low flag
Provides a deterministic maintenance signal and helps correlate time-loss complaints to low-voltage conditions.
Switchover detect
Counts how often backup events occur and detects unstable primary power or intermittent connections.
Optional tamper
Treat as a simple flag for audit purposes without expanding into security architecture.
Event logging fields (practical, production-friendly)
Power-fail TS Switchover count Battery-low duration Min VBAT Reset reason flag

Each field must map to an action: maintenance scheduling, battery aging judgment, switchover instability detection, and rapid triage of “lost time” complaints.

Low-power monitoring strategies (monitoring must be budgeted)
Duty-cycled sampling
Sample VBAT periodically or on events, not continuously. Reduce average load by design.
Gated divider
Keep the divider off most of the time. Enable briefly for measurement, then fully disconnect to avoid steady leakage.
Measurement isolation
Treat ADC input networks, clamps, and test fixtures as leakage contributors. Validate at hot and after soak.
Minimal verification points (observability that does not drain VBAT)
  • Verify gated divider off-state leakage fits the budget; on-state measurement completes within a short window.
  • Confirm event timestamps and counters are consistent with repeated VCC removal/re-apply cycles.
  • Repeat at room and hot; monitoring network leakage that grows at hot must be treated as worst-case.
Figure 8 — Observability blocks (VBAT monitor + gated divider + event logger)
Observability Blocks for RTC Backup Block diagram showing VBAT monitored through a gated divider into an ADC, with battery-low and switchover flags recorded in an event logger read by the host when awake. VBAT Gated divider Gate ADC Flags battery-low switchover Event logger power-fail TS switchover count battery-low duration Host / MCU when awake Monitoring must be budgeted gated + duty-cycled

PCB & Mechanical Design for Backup Domain (Cleanliness, Guarding, and Return Paths)

Ultra-low leakage is often decided by board implementation rather than IC datasheet current. The backup domain must be treated as a clean, short, isolated island with controlled return paths and verified hot/humidity behavior.

Backup island implementation rules (layout + assembly)
Partition
Keep VBAT routing short and local. Maintain a keepout band to separate the backup island from high dv/dt and high-current regions.
Return paths
Provide a clean, short reference return for the backup island. Avoid routing backup returns across switching current loops that can inject ground bounce.
Cleanliness
“No-clean” does not mean “no leakage.” Define a critical clean zone around VBAT, battery holder, and test points; validate after humidity/soak.
Component & mechanical traps (common leakage dominators)
ESD / TVS
Reverse leakage at hot can dominate the entire budget. Always evaluate temperature dependence for parts near VBAT.
High-value resistors
High impedance nodes become surface-leakage sensitive. Verify behavior after cleaning, humidity exposure, and handling.
Battery holder / connector
Contact contamination and intermittent resistance can cause micro-dips and unpredictable leakage paths. Mechanical retention and assembly hygiene matter.
Minimal practical checks (what must be true on real boards)
  • VBAT routing is short, direct, and separated from switching nodes by a clear keepout band.
  • Backup island return path is short and does not cross high-current switching loops.
  • Critical clean zone is defined around VBAT island, battery holder, and measurement points; verified after soak/humidity exposure.
  • ESD/TVS parts near VBAT are reviewed for hot reverse leakage as part of the leakage budget.
Figure 9 — PCB “backup island” layout concept (partition, guard, keepout, return)
PCB Backup Island Layout Concept Abstract board diagram with a VBAT island, guard ring, keepout zone, short VBAT route, correct return path, and a high dv/dt zone marked as avoid. High dv/dt zone SW / gate drive SW node VBAT island Keepout ORing Battery holder clean zone RTC VBAT backup pin Guard ring Long route Clean return No cross Do Short Clean Isolate Don’t SW Hot leak Long trace

Validation & Production Test (Measure Leakage and Holdover Without Fooling Yourself)

Leakage and holdover measurements are easy to falsify with instrument burden voltage, contaminated fixtures, and hot leakage drift. A reusable test flow must separate measurement artifacts from real budget failures.

Common traps (why “good” data can be wrong)
Burden voltage
Current-range burden can pull VBAT off its real operating point and distort the measured Ibackup.
Fixture contamination
Probes, cables, and dirty test points add parallel leakage. Handling and humidity can shift readings in nA–µA ranges.
Hot leakage drift
ESD/TVS and clamp devices can look fine at room and fail budget at hot. Temperature must be part of the validation plan.
Reusable validation flow (lab + production compatible)
Step 1: Baseline Ibackup
Measure at the intended VBAT operating point. Stabilize handling, cleanliness, and environment before recording baseline.
Step 2: Cut VCC
Use a repeatable cut method (relay or electronic switch). Avoid creating an unintended discharge path through instrumentation.
Step 3: Observe
Capture VBAT during the transition and confirm time/flags remain consistent. Log switchover events without adding steady leakage.
Step 4: Pass/Fail
Apply a single production criterion and confirm post-event integrity (no reset, expected flags, stable VBAT behavior).
Production simplification (2–3 points + 1 criterion)
TP1: Ibackup
Record the steady backup-domain current at a defined VBAT condition; treat fixture cleanliness as part of the test definition.
TP2: VBAT transition
Spot-check VBAT dips during VCC cut/re-apply on sampling basis; correlate with reset/time-integrity outcomes.
TP3: Flag integrity
Confirm expected power-fail/switchover flags and counters are updated and do not indicate unintended reset events.
Single criterion template
Ibackup < X µA @ 25°C (with periodic hot audit) and post-switchover flags indicate normal integrity.
Figure 10 — Test flow (baseline → cut VCC → observe → pass/fail)
RTC Backup Validation Test Flow Four-step flow diagram with icons and minimal labels: measure Ibackup, cut VCC, observe VBAT and flags, then decide pass or fail; includes small warning tags for measurement traps. Step 1 Step 2 Step 3 Step 4 Baseline Cut VCC Observe Decision Ibackup VBAT + flag Trap tags Burden Fixture Touch Hot leakage drift Repeatable edge Single criterion

Applications & IC Selection Notes (Strictly Within This Page’s Boundary)

Selection is constrained to backup power, switchover integrity, and ultra-low leakage. Avoid solutions that add steady µA-class drain or create any coin-cell charge path.

Scenario buckets (decision drivers only)
Industrial instrumentation
Long holdover, low maintenance. Priority is minimizing hot/humidity leakage and preventing event-loss on switchover.
Typical direction
Coin-cell or Hybrid (coin-cell for long-term + supercap for transients), with gated monitoring only.
Gateways / edge devices
Maintainability and observability matter, but monitoring must not become the dominant leakage path.
Typical direction
Built-in switchover or ideal-diode OR, plus event flags/logs and duty-cycled battery measurements.
Automotive / low temperature
Low-temperature impedance and vibration drive micro-dip risk; switchover hysteresis and mechanical contacts become critical.
Typical direction
Hybrid with a controlled supercap charge path and verified reverse isolation; robust holder/connector strategy.
Selection fields (only what affects backup/switchover/leakage)
VBAT operating range
Confirm the guaranteed region across temperature and during transient switchover dips.
Ibackup vs temperature
Use worst-case curves; room-temperature “typical” is not a holdover guarantee.
Switchover threshold & hysteresis
Prevent chatter under slow ramps/noise and ensure no reset/time-loss during the transition.
Reverse leakage / backfeed limits
Validate reverse paths at hot; avoid any VBAT → VCC or VCC → VBAT unintended conduction.
Charge path features (supercap)
Require current limit, over-voltage clamp, charge termination/indication (if needed), and reverse isolation.
Flags / logging hooks
Prefer event flags/counters, and implement battery measurement with gated divider or duty cycling to avoid steady drain.
Prohibited items (to avoid µA-class “hidden” drain and safety hazards)
  • Always-on divider / always-on ADC monitoring that adds steady µA drain.
  • Any coin-cell “trickle charge” path, including accidental backfeed through protection or diode paths.
  • “More TVS near VBAT” without verifying hot reverse leakage and its impact on holdover.
Concrete material numbers (starting points only)

The items below are examples to speed up datasheet lookup and lab verification. Always verify package, suffix, voltage rating, and temperature leakage in the target operating range.

Boundary reminder
Only backup/switchover/leakage-related choices are listed. No RTC register mapping, no secure-time design.
RTC IC examples (VBAT pin + backup domain)
  • Microchip MCP79410 (battery backup capable RTC family)
  • Microchip MCP7940N (backup-capable RTC family)
  • NXP PCF85263A (RTC with backup domain capability)
  • NXP PCF8563 / PCF85063A families (backup supply pin variants; confirm exact suffix)
What to verify
VBAT range, Ibackup vs temperature, power-fail flags/event logging hooks, and switchover behavior.
Switchover / power-path examples
  • Analog Devices/Linear Tech LTC4412 (ideal diode controller family)
  • Analog Devices/Linear Tech LTC4413 (power-path controller family)
  • Maxim/Analog Devices MAX40200 (ideal diode / ORing style device; confirm leakage at hot)
  • TI TPS2113A / TPS2121 (power mux class; check quiescent and reverse blocking carefully)
What to verify
Reverse leakage, backfeed blocking behavior, static IQ, and switchover dip/glitch sensitivity.
Discrete ORing (PFET / Schottky) examples
  • PFET: AO3407A (SOT-23 PFET class)
  • PFET: SI2301 / SI2305 class parts (confirm polarity and leakage)
  • Schottky: BAT54 family (tiny Schottky; verify hot IR)
  • Schottky: BAS40 family (low-leakage Schottky class; verify hot IR)
What to verify
PFET body-diode paths and gate biasing; Schottky reverse leakage at temperature and the resulting VBAT drop.
Energy source & holder examples
  • Coin-cell: Panasonic CR2032 (verify exact series and tabs)
  • Coin-cell holder: Keystone 3000 / 3008 series (CR2032 holders; verify mounting style)
  • Supercap family: Panasonic Gold Cap (EEC series; choose voltage rating with margin)
  • Supercap family: AVX/Kyocera BestCap (SC series; verify leakage and ESR)
What to verify
Coin-cell: no-charge guarantee & low-temperature impedance. Supercap: leakage vs temperature and a safe charge clamp strategy.
Gated monitoring building blocks (avoid steady µA drain)
  • Small-signal NMOS for divider gating: 2N7002 (verify leakage class and Vgs headroom)
  • Small-signal NMOS option: BSS138 (verify leakage and Vds rating)
  • Load switch class: TI TPS22910A / TPS22916 (verify off-leakage and on-resistance)
  • ESD diode family (board entry, not core VBAT node): Nexperia PESD5V0 family (verify hot leakage)
What to verify
Off-state leakage and hot leakage dominate; ensure monitoring is duty-cycled and the divider is physically in the clean backup island.
Figure 11 — Selection funnel (inputs → energy source → topology)
Selection Funnel for RTC Backup & Switchover Inputs feed a funnel that selects coin-cell, supercap, or hybrid, then chooses topology: built-in, diode OR, ideal OR, or always-on island. Holdover target Temperature Maintenance Energy choice Coin-cell Low leak Supercap Pulse Hybrid Balance Topology choice Built-in Check backfeed Diode OR Drop / hot IR Ideal OR Leak / IQ Island Clean layout No coin-cell charge No always-on divider Check hot leak

Engineering Checklist (Design → Layout → Test → Production)

This checklist converts the page into a repeatable execution plan. Each line is phrased as Check → Evidence → Pass to support reviews, lab work, and production gates.

Design (Spec lock)
  • Ibackup target defined at room + hot; evidence: budget sheet; pass: sum < target.
  • Holdover target defined with VBAT min; evidence: calculation + margin; pass: meets hours/days goal.
  • Switchover integrity criteria fixed; evidence: test plan; pass: no reset/time-loss + correct flags.
  • Monitoring allowed rules; evidence: gated/duty-cycled scheme; pass: no steady µA drain.
Schematic review
  • Backfeed paths enumerated; evidence: current-flow markup; pass: no coin-cell charge path.
  • Protection parts reviewed for hot leakage; evidence: datasheet worst-case; pass: budget OK at hot.
  • Divider/ADC measurement gated; evidence: switch schematic; pass: off-leak dominates < budget.
  • Supercap charge has limit + clamp + reverse isolation; evidence: block review; pass: safe and low-leak.
Layout & process
  • VBAT island and keepout implemented; evidence: layout screenshot; pass: away from SW node.
  • Clean return path; evidence: return-path check; pass: no crossing of high-current loops.
  • Critical clean zone defined; evidence: assembly note; pass: soak/humidity re-test required.
  • Battery holder/connector stability; evidence: mechanical retention; pass: no micro-dip in shake test.
Lab validation
  • Ibackup measured at defined VBAT; evidence: setup photo + reading; pass: within budget.
  • VCC cut method repeatable; evidence: waveform; pass: consistent edge & no unintended discharge path.
  • Switchover dip captured; evidence: VBAT + flag log; pass: no reset/time-loss.
  • Hot/humidity audit; evidence: before/after delta; pass: leakage remains within margin.
Production
  • TP1 Ibackup defined; evidence: fixture spec; pass: < X µA @ 25°C.
  • TP3 flag integrity checked; evidence: readout; pass: expected power-fail/switchover indicators.
  • Hot audit plan defined; evidence: sampling rate; pass: drift stays within guardband.
  • Failure attribution flow; evidence: isolate steps; pass: board leak vs device leak identified.
Reference BOM snippets by block (examples to accelerate builds)

These examples map to the checklist blocks above. Verify leakage at hot and confirm exact suffix/package for procurement.

Switchover controller / mux
LTC4412, LTC4413, TPS2113A, TPS2121
Discrete ORing
AO3407A (PFET), 2N7002 (gating), BAT54 (Schottky), BAS40 (Schottky)
Energy source & holder
CR2032 (coin-cell), Keystone 3000/3008 (holders), Panasonic Gold Cap (EEC series), AVX BestCap (SC series)
RTC IC family
MCP79410, MCP7940N, PCF85263A, PCF8563 / PCF85063A families
Figure 12 — Checklist pipeline (Design → Layout → Validate → Production)
Engineering Checklist Pipeline Four-stage pipeline showing Design, Layout, Validate, and Production blocks with three checkboxes per stage. Design Layout Validate Production Spec Backfeed Monitor Island Clean Return Ibackup Cut VCC Flags

Request a Quote

Accepted Formats

pdf, csv, xls, xlsx, zip

Attachment

Drag & drop files here or use the button below.

FAQs (Backup Current, Holdover, Switchover Integrity)

These FAQs are strictly scoped to backup leakage, energy holdover, switchover glitches/backfeed, monitoring without drain, and measurement/production screening. Each answer is short, actionable, and includes pass/fail criteria.

Holdover is 10× shorter than calculated—what’s the first leakage isolation step?
Likely cause: One “non-modeled” path dominates (board contamination, protection leakage, divider/monitor drain, or reverse path through ORing).
Quick check: Remove power-step variables—force VBAT-only mode, then isolate by (1) opening the monitor divider, (2) temporarily lifting/leaving out the ESD/TVS on VBAT, (3) disconnecting external cables/connectors from the backup island.
Fix: Clean/coat the backup island, gate the divider, move/replace high-leak protection parts, and eliminate reverse conduction paths (correct PFET orientation / ideal diode wiring / body-diode paths).
Pass criteria: Ibackup_total ≤ X (set X to your budget target) and the measured holdover matches the model within ±20% at the same VBAT and temperature.
VBAT current looks fine at 25°C but fails at 60°C—what parts to suspect first?
Likely cause: A temperature-accelerated leakage device dominates (Schottky IR, TVS/ESD reverse leakage, PFET/ideal-diode controller leakage, or supercap self-leak/charge clamp leak).
Quick check: Heat only the suspect area (hot air/thermal pad) while logging Ibackup; then swap one component at a time: Schottky → alternative, TVS/ESD → alternative, or temporarily remove the clamp/monitor chain to see the step change.
Fix: Use a lower-leakage protection device, avoid Schottky ORing where hot IR breaks the budget, and gate any monitor divider; for supercap, validate leakage vs temperature and adjust clamp/topology.
Pass criteria: Ibackup_total ≤ Y at hot (Y = hot-budget target) with ΔI(25→60°C) consistent with the modeled worst-case; no single part change causes >50% budget swing.
Why does swapping the ESD/TVS part change backup life dramatically?
Likely cause: Reverse leakage (IR) varies widely by TVS/ESD family, especially at elevated temperature; placement can also create unintended bias conditions that increase leakage.
Quick check: Measure Ibackup with the TVS/ESD installed vs removed (or replaced with a known-low-leak part) at the same VBAT and temperature; repeat at hot to expose leakage acceleration.
Fix: Select protection parts by hot IR specification (not just room typical), and avoid placing high-leak clamps directly on the most sensitive VBAT node; protect at less sensitive nodes if possible.
Pass criteria: TVS/ESD contribution to Ibackup is ≤ 10–20% of the total budget at hot, and the holdover delta after substitution stays within ±10%.
Switchover “works” but RTC resets on fast VCC drop—what waveform should I capture first?
Likely cause: A brief VBAT dip/glitch crosses the RTC backup threshold, often from ORing transition, ground bounce, long lead inductance, or backfeed clamp behavior.
Quick check: Capture 4 traces on the same timebase: VCC, VBAT at the RTC pin, the ORing/switchover node (after diode/PFET/ideal diode), and a close ground reference (short ground spring). Trigger on VCC falling edge and zoom into the first 1–10 ms.
Fix: Reduce path impedance (short VBAT route, local bulk/decoupling on VBAT), add hysteresis or improve ORing control, remove or relocate high-leak clamps that distort transitions, and ensure the backup island ground return is clean.
Pass criteria: VBAT_min_dip ≥ VBAT_min_spec + margin (typical margin: 50–150 mV) and no reset/time-loss flags across repeated fast-drop tests (≥50 cycles).
Coin-cell holds time, but supercap solution doesn’t—what’s the top self-leak check?
Likely cause: Supercap self-leakage (and/or the charge clamp path) dominates; the energy looks large on paper but leaks away faster than the load consumes it.
Quick check: Disconnect the RTC load and measure supercap discharge rate alone (V(t)) at a fixed temperature; separately measure I into the clamp/charger path with VCC removed to expose “hidden” drain.
Fix: Choose a lower-leakage supercap family, redesign the clamp to eliminate reverse or bias leakage, and avoid any always-biased monitoring that keeps the cap bleeding.
Pass criteria: With load removed, the equivalent self-leak current is ≤ 0.3× the allowed Ibackup budget; with load connected, measured holdover matches the capacitor-based model within ±20%.
Why does adding a VBAT divider for ADC kill battery life—how to gate it correctly?
Likely cause: The divider creates a permanent DC load: I = VBAT / (Rtop + Rbot), which can exceed the RTC’s own backup current by orders of magnitude.
Quick check: Compute divider current and compare to Ibackup budget; then temporarily open the divider and re-measure Ibackup_total to confirm the step change.
Fix: Gate the divider with a MOSFET/load-switch controlled by a timer/event; sample briefly (e.g., 1–10 ms), then fully disconnect. Keep the divider physically inside the cleaned backup island.
Pass criteria: Average monitoring drain Iavg_monitor ≤ 0.1× Ibackup budget (target), and gated “OFF” leakage stays below (X − I_RTC − margin).
Backup is stable on bench but fails in humid environment—what board-level leakage test is fastest?
Likely cause: Surface conduction from flux residue, contamination, or hygroscopic materials creates parallel leakage across high-impedance nodes in the backup island.
Quick check: Run a humidity/soak A/B: measure Ibackup before soak, after soak, then after targeted cleaning (IPA + controlled process) on the VBAT island. A large “cleaning recovery” indicates board leakage dominance.
Fix: Define a critical clean zone, enforce cleaning/handling rules, add keepouts/guarding where appropriate, and consider conformal coating only after leakage baseline is proven.
Pass criteria: After soak, ΔIbackup ≤ ΔI_limit (set by guardband) and post-cleaning does not change Ibackup by more than 10–15%.
VBAT backfeeds into VCC rail—what’s the most common schematic mistake?
Likely cause: Unblocked reverse path through a PFET body diode, diode-OR orientation error, power-mux reverse blocking not guaranteed, or a clamp/ESD path to VCC.
Quick check: With VCC removed and VBAT present, measure VCC rise and current into VCC; then isolate by disconnecting the ORing element and any clamps tied to VCC to identify the reverse path.
Fix: Correct PFET orientation and gate biasing, use an ideal-diode controller with verified reverse blocking, and keep high-leak clamps off the VBAT core node (or select lower-leak alternatives).
Pass criteria: With VBAT applied and VCC off, VCC ≤ 0.1 V (or per system spec) and reverse current into VCC is ≤ I_rev_limit (set to a small fraction of Ibackup budget).
After battery replacement, time is wrong—how to design a “safe replace” flow without extra leakage?
Likely cause: VBAT drops below the backup retention threshold during the swap, or the “replace” procedure triggers repeated switchover dips; no reliable event marker exists to detect/repair time loss.
Quick check: Measure VBAT at the RTC pin during a real swap (fast + slow) and log power-fail/switchover indicators; confirm whether VBAT crosses the RTC’s minimum retention threshold.
Fix: Provide a service-mode hold-up path (e.g., temporary external VBAT via test pads, or a small hold-up capacitor/supercap used only during swap) plus an event marker (battery-replace flag) to force time resync when needed—without a permanent always-on load.
Pass criteria: During swap, VBAT_min ≥ VBAT_min_spec + margin and the system logs a replace event; if VBAT dips, the system reliably triggers a resync path within T_resync.
Why does a higher-value resistor still leak more than expected—what non-ideal path is likely?
Likely cause: The leakage is not through the resistor value; it is via surface conduction, solder mask/flux residue, ESD clamp leakage, connector contamination, or measurement fixture leakage in parallel.
Quick check: Lift one end of the resistor to break the intended path; if Ibackup remains high, the current is flowing elsewhere. Then clean the area and re-test; a large delta indicates board leakage.
Fix: Reduce exposed high-impedance surface area, move/gate the divider, define a clean zone, and replace/relocate protection parts that leak at hot.
Pass criteria: With the divider opened, Ibackup drops to near the RTC+switch baseline, and cleaning changes Ibackup by ≤ 10–15%.
How do I measure nA-level backup current without the meter corrupting the circuit?
Likely cause: Series ammeters add burden voltage and range switching artifacts; cabling/fixture contamination or guardless probing creates parallel leakage that dominates nA readings.
Quick check: Use a known VBAT and measure current via voltage drop across a precision sense resistor (DMM in voltage mode) or use an SMU/picoammeter; repeat with short, clean connections and a guarded setup if available.
Fix: Avoid series DMM current mode for nA; prefer SMU/current-measurement supply, or a calibrated sense resistor method; keep the backup island clean and eliminate long, humid cabling during measurement.
Pass criteria: Two methods agree within ±10–20% and VBAT at the RTC pin stays within ±10 mV of the intended test point during measurement.
What is a reasonable pass criteria set for production screening (Ibackup + switchover)?
Likely cause: Production failures come from board leakage variance, protection part leakage spread, or marginal switchover dips not caught by a single “room-only” current check.
Quick check: Use a 2–3 point screen: (1) Ibackup at room, (2) a fast VCC cut to verify switchover integrity flags, and (3) periodic hot/humidity audit sampling to catch drift/leakage acceleration.
Fix: Set guardbands from budget + process spread; define a clean/handling process; add a failure attribution flow (board leak vs device leak) for rapid containment.
Pass criteria: Ibackup_room < X, switchover test shows no reset/time-loss + expected flags, and audit results keep P99 Ibackup_hot < Y (X/Y = budget-based limits with margin).
icnavigator

ICNavigator helps engineers find verified ICs, alternatives, and fast quotes.

Explore
Categories
  • Power Management
  • MCU
  • Automotive ICs
  • Sensors
  • Interface & Transceivers
  • Analog Front-End
Get in Touch
  • Email: hello@icnavigator.com
  • WeChat/WhatsApp: +86-xxxx

© 2025 ICNavigator. All rights reserved.