YURUNOX / Low-power component selection

What Is Quiescent Current and Why Does It Matter?

Quiescent current, or IQ, is the current an IC uses to keep its internal circuitry operating under specified idle conditions. Because the value depends on voltage, load, mode, temperature and measurement boundary, compare only equivalent conditions and then validate total current at the battery terminals.

The practical target is not automatically the smallest datasheet figure. Identify the function that must stay awake, assign a current budget to every power state, and confirm that lower idle consumption does not break startup, regulation, noise or timing requirements.

By YURUNOX · For engineers and technical component buyers
Source review:

CR2032 coin-cell battery showing its compact metal case
Battery current across active and sleep statesA battery supplies short active-current pulses and a continuous lower sleep current. Both contributions determine the average current.BATactive pulses + continuous sleep currentaverage current sets the ideal runtime model

Fallback diagram: calculate active and sleep contributions at the same battery connection.

A small battery makes continuous background consumption important. The worked example below uses an assumed usable capacity, not the capacity of this photographed cell. Photo: JanEhlebrecht, Wikimedia Commons, CC BY-SA 4.0. No further edits.

What does quiescent current actually measure?

It measures the current an IC consumes in a specifically defined operating condition. An enabled regulator may still run its reference and control circuit. An amplifier may still bias its signal path. A wake detector must remain ready to notice an event.

IQ describes one component under stated conditions; system sleep current describes the complete product in a stated operating mode. They are not automatically the same measurement.

Swipe the table sideways to compare all columns.

Match the operating state to the evidence you need
ConditionRecommendationEvidence requiredStop boundary
Always-on wake or retention functionBudget the enabled-state current.IQ limits at the required voltage, load, mode and temperature.Do not substitute shutdown current for a function that must remain active.
Periodically active battery productCalculate the complete cycle at the battery connection.Sleep current, active waveform, duration, frequency and usable capacity.Do not claim runtime from one IC's IQ alone.
Regulator candidate comparisonCompare equivalent no-load and loaded operating points.Mode pins, no-load input current, efficiency curves, startup and load-step data.Stop if the low-current mode is unavailable in the selected package or configuration.
Disabled storage or shipping stateTrace every path still connected to the battery.Shutdown limits plus charger, protection, pull-up and leakage paths.Do not treat a disabled main rail as zero system current.

A useful first question is: what does this current keep alive? If that function must respond continuously, turning it off changes the product. If it supplies an unused rail, shutdown may be worth evaluating.

Exposed integrated-circuit die connected to package leads by bond wires
Internal IC functions that consume quiescent currentA supply feeds reference, bias and control blocks before the device output, showing why an enabled IC can consume current without an external load.SUPPLYreferencebiascontroloutputenabled internal functions can draw current at zero external load

Fallback diagram: the specification must state which internal functions and modes remain active.

Inside an IC, supply and signal connections serve different current paths. Package appearance cannot establish its IQ.Photo: Mister rf, Wikimedia Commons, CC BY-SA 4.0. No further edits.

Which definition and test condition sit behind the IQ symbol?

For a switching converter, a nonswitching IQ test can isolate internal bias consumption. Real no-load operation may include occasional pulses that maintain the output voltage. Both numbers are useful, but they answer different questions.

Some devices also draw operating current from more than one supply pin. Do not add currents from different voltage rails and call the result battery current.

Technical basis: TI's explanation of three IQ specifications.

Unit check: 1 mA = 1,000 µA; 1 µA = 1,000 nA. A missing unit conversion can change a battery estimate by a factor of 1,000.

02 / Define the measurement boundary

Which current specification belongs in your power budget?

Compare the parameter description and test conditions, not just the symbol. Manufacturers do not use every current label identically.

Swipe the table sideways to compare all columns.

Which current belongs in your calculation?
TermWhat it describesQuestion to ask
Quiescent current, IQInternal operating consumption at a defined point.Enabled? Switching? Loaded? Which supply pins?
Shutdown current, ISDConsumption in a disabled state.What still works, and what other paths remain connected?
Ground current, IGNDCurrent returning through an LDO's ground connection.At what output load and input voltage?
No-load input currentInput consumption while maintaining the output with no external load.Which conversion mode and voltage ratio?
System sleep currentTotal board consumption in a defined firmware and hardware state.Was it measured at the battery connection?
Input bias currentCurrent at an amplifier's signal-input terminals.Is it being confused with supply consumption?

How do you trace current from the battery into an LDO?

For a simple single-input LDO in steady state, input current is approximately output current plus ground current. At zero output load, input and ground current are approximately equal. With a real load connected, they are not.

Battery current includes several branchesBattery current splits between an LDO input and an always-on branch. The LDO input splits into load current and ground current. This simplified steady-state LDO diagram is not a buck-converter current model. BatteryI BAT I IN LDO I OUT Sleepingload Always-on branchI AUX I GND I BAT = I OUT + I GND + I AUX
Original explanatory diagram: simplified LDO current accounting, with return wiring omitted. Include any separate bias supplies or additional branches. Do not apply this current-sum model directly to a switching converter.

For the LDO relationship and its limits, see TI, Understanding the Terms and Definitions of LDO Voltage Regulators.

03 / Make the impact tangible

How much can a few microamps change battery life?

An additional continuous battery-side drain of 20 µA consumes 175.2 mAh in 8,760 hours. This is simply charge = current × time, not a prediction for a particular cell.

Illustrative engineering scenario · not a measured customer result

What changes when a sensor wakes for 0.1 second every minute?

Assume 500 mAh of usable battery capacity. The complete active event draws 10 mA for 0.1 s, including startup, measurement and transmission. The remaining 59.9 s draws either 30 µA or 10 µA.

These are total battery-terminal currents. Regulator overhead, enabled support circuits and conversion losses are already included; they must not be added again.

IAVG = (IACTIVE × tACTIVE + ISLEEP × tSLEEP) / tCYCLE
Same active event, lower sleep current
Budget itemDesign A: 30 µA sleepDesign B: 10 µA sleep
Active contribution16.67 µA average16.67 µA average
Sleep contribution29.95 µA average9.98 µA average
Total average46.62 µA26.65 µA
Ideal calculated runtime10,726 hours / 447 days18,762 hours / 782 days
Where the average current goes in this example
Active-event contribution Sleep contribution
Design A46.62 µA
Design B26.65 µA

Both bars use the same scale. Calculated contributions, not an oscilloscope trace.

The model gives approximately 75% longer runtime for Design B. It does not establish that changing one IC will achieve this result. The full board must actually deliver the assumed sleep-current reduction.

There is also a useful counterexample: at 50% active time, the same 10 mA active level contributes 5 mA to the average. Saving 20 µA only during sleep then saves 10 µA overall—about 0.2% of the original average. In that product, active power or wake duration may deserve attention first.

04 / Use one consistent measurement boundary

How do you calculate average battery current and runtime?

Compare two sleep-current targets while keeping the active event unchanged. Enter currents measured or estimated at the same battery connection, not a mix of output-rail currents.

Design A average
46.62 µA
Design B average
26.65 µA
Design A ideal runtime
447 days
Design B ideal runtime
782 days

Design B has 75% longer ideal runtime than Design A.

Active share: 0.167%. Sleep time: 59.9 seconds per cycle.

Planning model only. Runtime (hours) = usable capacity (mAh) / average battery current (mA). Each state is represented by its mean current. Include every wake transition in the active event. Battery self-discharge, aging, temperature effects, pulse limits, changing conversion efficiency and cutoff behavior are not modeled.

For a production target, use justified usable capacity and measured duty cycles. A long mathematical runtime is not evidence that a battery will support that many years in storage or service.

05 / Separate idle overhead from conversion loss

Why can a low-IQ regulator still waste power?

A regulator can have very low idle consumption and still be the wrong architecture for the active load. Separate idle overhead from voltage-conversion loss.

How do voltage drop and ground current combine in an LDO?

η = (VOUT × IOUT) / [VIN × (IOUT + IGND)]

Using the simplified single-input model, take 3.6 V in, 3.0 V out and a 10 µA load. If ground current at that operating point is 2 µA, efficiency is about 69.4%. At 20 µA ground current, it is about 27.8%. These are illustrative values, not measurements of a named regulator.

Even negligible ground current would leave an ideal voltage-ratio limit of 3.0 / 3.6, or 83.3%. At heavier loads, the voltage drop can matter more than the idle-current headline. TI's LDO definitions report gives the underlying current and efficiency relationships.

How should buck-converter input current be estimated?

IIN = (VOUT × IOUT) / (VIN × η)

Use efficiency as a decimal at the required input voltage, output voltage, load and operating mode. If the efficiency already includes internal losses, do not add IQ again. At zero load, use the specified or measured no-load input current instead.

A low-IQ search filter can identify candidates. It cannot replace light-load efficiency curves, a complete power budget or testing during the largest load step.

06 / Compare like-for-like conditions

What must match before you compare quiescent-current datasheets?

These devices illustrate how to interpret specifications. They perform different functions and are not drop-in alternatives or a ranking. Values below are taken from the linked datasheet revisions.

Published specification / TI TPS7A02 LDO

What does TPS7A02 show about typical and maximum IQ?

The no-load ground-current row gives 25 nA typical and 46 nA maximum at 25°C. The maximum becomes 60 nA over −40°C to +85°C. That row does not extend the 60 nA limit to +125°C.

Conditions include zero output load, EN tied to VIN, 1 µF input/output capacitors and VIN equal to the greater of VOUT(nom) + 0.5 V or 2.0 V. Shutdown is separate: 3 nA typical at 25°C, EN = 0 V and VIN = 1.5–5.0 V.

Selection consequence: an always-on rail needs an enabled-state budget. A strict hot-temperature limit needs evidence covering that temperature, not a room-temperature headline.

TPS7A02 datasheet, Rev. C, section 6.5.

Published specification / TI TPS62840 buck converter

What does TPS62840 show about mode-dependent no-load current?

At 3.6 V input, 1.8 V output, zero load and 25°C, the datasheet lists 60 nA typical in power-save operation with EN = VIN and MODE/STOP grounded. The device is switching as needed, not externally held in a nonswitching test state.

Power-save no-load operation60 nA typical
Forced PWM no-load operation3 mA typical

For the stated 1.8 V output condition, forcing MODE high changes the typical no-load input-current figure to 3 mA. Separate nonswitching VIN and VOS current rows describe another measurement.

Selection consequence: review the chosen package's available mode pins and actual mode setting before assuming the advertised low-current operation applies.

TPS62840 datasheet, Rev. D, sections 5 and 7.5.

Published specification / TI OPA391DCK amplifier

What does OPA391 show about supply current versus input bias current?

The quiescent-current row specifies 23.5 µA typical per amplifier and 30 µA maximum at 25°C. Its test condition sets input common-mode voltage to (V+) − 1.5 V. Default table conditions include a 1.7–5.5 V supply and a 10 kΩ load.

Selection consequence: a tiny input-bias-current specification is not the amplifier's supply demand. For multi-channel candidates, count enabled channels and include the external signal network in the board budget.

OPAx391 datasheet, Rev. E, section 5.7.

When a relevant maximum is missing, record that gap. Typical curves help estimation, but they do not become guaranteed acceptance limits. Keep the full ordering code and datasheet revision with the comparison.

07 / Preserve the required function

Should the circuit sleep, stay enabled or shut down?

  1. ALWAYS-ON RAILKeep the required watcher awakeA wake detector or retention circuit may need continuous power. Optimize its enabled-state overhead.
  2. INTERMITTENT RAILCompare off time with restart costAccount for rail charging, initialization and settling before choosing power gating.
  3. STORAGE STATEBudget the months before useCount every protection, charger or wake circuit still connected while the product is disabled.
Illustrative break-even calculation

How long must the off interval be before shutdown saves charge?

Suppose shutting down saves 20 µA at the battery, but each restart consumes an extra 200 µC compared with leaving the rail enabled. The break-even off time is 200 µC / 20 µA = 10 seconds.

At 2 seconds off, the saved charge is only 40 µC—less than the restart penalty. At 60 seconds off, it is 1,200 µC, before subtracting that penalty. This simplified comparison assumes the useful work is identical and the restart charge is incremental.

Energy is only one constraint. The circuit must still wake in time, preserve necessary state and settle before sampling. For regulators, check startup, load-step response, dropout, capacitor requirements, noise and power-supply rejection. Noise and ripple rejection are different specifications.

Power-saving switching modes can alter ripple patterns without indicating instability. Conversely, do not assume every regulator permits zero load; check its specific minimum-load requirement. See Analog Devices' linear-regulator selection guidance and zero-load operation discussion.

08 / Reproduce the operating state on the bench

How should you measure quiescent and standby current?

Choose the test question first: Does one IC match a specified condition? Does the board meet a sleep target? What charge does a complete cycle consume? Each needs a different measurement boundary.

Fluke digital multimeter with rotary range selector and input sockets
Low-current measurement with voltage burdenA battery, ammeter and device under test are connected in series. The diagram highlights the voltage drop across the ammeter and the need to verify voltage at the device.BATTERYAMMETERburden voltageDEVICEunder testverify the voltage and mode at the device pins

Fallback diagram: the measuring instrument can change the circuit it is measuring.

Instrument range, burden voltage and resolution matter. This equipment photo is not a recommended nanoamp test setup.Photo: Alex P. Kok, Wikimedia Commons, CC BY-SA 4.0. No further edits.

How can the meter change the operating point?

A current meter introduces a voltage drop, called burden voltage. Check voltage at the device pins, including during wakeup. A range that resolves sleep current may not tolerate the active peak without disturbing the rail.

At nanoamp levels, offset, fixture leakage, humidity, contamination and settling time can affect the result. More display digits do not necessarily mean adequate accuracy.

Measurement reference: Keithley Low Level Measurements Handbook, section 2.3.

  1. Reproduce the device conditionRecord voltage, temperature, output setting, load and control-pin levels. For a no-load test, remove external loads while retaining required support components.
  2. Remove unintended contributorsCheck evaluation-board LEDs, debugger connections and interface circuits. Disconnect development accessories safely and document the test configuration.
  3. Capture a complete operating cycleMeasure at the battery connection for runtime work. Use a setup that resolves sleep current and captures active peaks, then integrate charge over representative cycles.
  4. Repeat at meaningful cornersCheck relevant battery voltages and temperatures, with realistic firmware and wake sources. A single quiet room-temperature reading does not validate the whole product.

Do not force a converter output externally merely to stop switching unless the manufacturer's test method permits it. Reverse-current paths or a changed bias supply can invalidate the measurement.

Why can board standby current exceed one IC's IQ?

Trace the extra current by branch and operating state
SuspectPractical checkWhat it can reveal
Divider or pull-upCalculate current using actual voltage and resistance.A 1 MΩ total divider across 3.3 V draws 3.3 µA.
Wrong modeMeasure enable/mode pins and confirm firmware settings.The intended low-current state may never be entered.
Live signal into an unpowered ICReview pin limits and power sequencing.A signal path may feed the supposedly disabled rail.
Retries or periodic wakeupsRecord a longer current trace with firmware events.Short bursts can dominate a quiet-looking average.
Development hardwareRepeat with unnecessary accessories removed.The fixture, not just the product, may be consuming power.

For perspective, that 3.3 µA divider consumes 28.9 mAh over 8,760 hours. It can outweigh an improvement of a few dozen nanoamps in one IC. Do not raise resistance blindly: input leakage, bias current, noise and settling time still constrain the circuit.

09 / Turn the power budget into an RFQ

What must engineers and buyers specify before ordering?

A request for a “low-IQ regulator” leaves too much unspecified. An alternative can improve the headline number while changing output accuracy, startup, package pinout or required capacitors.

  1. Define each power stateSet separate enabled-idle, active and disabled-storage budgets. Identify the functions that must remain available in each state.
  2. Specify electrical and timing limitsProvide battery range, output voltage, sleep/peak load, temperature range and wake timing. Distinguish typical design targets from guaranteed limits.
  3. Identify the exact partInclude manufacturer, full ordering code, voltage option, package, pinout and datasheet revision. Family names alone do not establish compatibility.
  4. Approve the complete replacementReview mode controls, stability, noise, load steps and power sequencing. Validate hardware before treating a proposed alternative as drop-in.

Keep electrical approval separate from stock, price and traceability review. Use the YURUNOX quality-assurance information alongside your qualification requirements; sourcing documentation does not substitute for system-level power validation.

Component sourcing / Clearer requirementsSend the operating states, not only the IQ target

Send the full part number, quantity and required date, plus the input/output voltages, load profile, temperature range and maximum acceptable idle current. State whether alternatives are allowed and which limits cannot change.

YURUNOX is an electronic-component sourcing partner. Keep final electrical approval with your engineering team.

10 / Evidence for the engineering decision

Which technical documents support these calculations and examples?

Manufacturer specifications support the device examples; the battery and break-even scenarios are illustrative calculations. No YURUNOX customer test or field result is claimed.

  1. Texas Instruments: 3 Quiescent-current (Iq) Specifications to UnderstandSSZT118 · Shutdown, nonswitching and switching current definitions.
  2. Texas Instruments: Understanding the Terms and Definitions of LDO Voltage RegulatorsSLVA079 · Ground current and LDO efficiency.
  3. TPS7A02 datasheetRev. C · Section 6.5: ground-current and shutdown limits.
  4. TPS62840 datasheetRev. D · Sections 5 and 7.5: package features and mode-dependent current.
  5. OPAx391 datasheetRev. E · Section 5.7: OPA391DCK current specifications.
  6. Analog Devices: Five Things You Should Know About Linear RegulatorsSelection factors beyond a low-IQ headline.
  7. Analog Devices: Minimum Load Current Operation—Zero-Load OperationDevice-specific minimum-load behavior.
  8. Keithley: Low Level Measurements Handbook, Seventh EditionSection 2.3 · Low-current measurement errors and techniques.

Use the latest datasheet for the exact ordering code before design approval. Image credits and licenses appear beneath each photograph.

Cart (0 items)