How to Test Power Adapter Inrush Current Safely

Power adapter inrush current is the short AC-input current pulse drawn when an adapter is first energized and its input capacitors begin to charge. The peak can be far above the adapter’s normal operating current, but peak amperes alone do not tell an OEM whether the event is acceptable. Input voltage, switch-on phase, source impedance, temperature, restart interval, pulse duration, upstream protection, and the number of adapters starting together all affect the decision.

The practical task is to define the operating scenario, obtain a conditioned waveform for the exact model, and verify the complete input path. That prevents a seemingly small adapter choice from becoming a nuisance-trip or startup problem during system testing.

What is power adapter inrush current?

Inrush current is an input-side event at turn-on. When AC reaches a switch-mode power adapter, the rectifier and input energy-storage circuit begin charging. Before the reservoir capacitor is charged, the input can present a much lower impedance than it does during normal operation. The resulting current pulse is brief, then the adapter settles to its steady operating pattern.

This explains why a 100 W adapter can show a startup peak that is many times its normal RMS input current without continuously drawing that peak. It also explains why two datasheet lines must not be compared as if they describe the same quantity:

• Normal input current describes sustained operation under stated line and load conditions.
• Inrush current describes a short turn-on event under a separate set of conditions.

The input circuit may use an NTC thermistor or another limiting method. An NTC has relatively high resistance when cold. Its resistance falls as it self-heats during operation, reducing continuous loss. Other designs can use a resistor with a bypass device or an active limiting circuit. The topology helps explain behavior, but an OEM usually gets a better result by specifying the required startup performance instead of dictating an internal circuit.

Why is a peak-current number not enough?

A line such as “inrush current: 60 A” is incomplete unless the conditions travel with it. The value might be a typical result rather than a guaranteed maximum. It may describe a cold start at room temperature, one input voltage, one phase angle, one output load, and one measurement definition. Another supplier may report a maximum at high line or include a different portion of the waveform.

Ask for the conditions behind the number:

These fields matter because the limiter, capacitor, rectifier, mains waveform, and source impedance interact. A high-line test may produce a different peak from a low-line test. Closing near an AC voltage peak can differ from closing near a zero crossing. A long extension lead, programmable AC source, UPS, or generator may present a different impedance from a stiff laboratory branch circuit.

Even “cold start” needs a defined initial condition. If the design uses an NTC limiter, a quick off-on cycle can leave the thermistor hot and at lower resistance. The restart event can therefore differ from the room-temperature cold-start result. The correct off-time is the one your product can experience, not an arbitrary pause selected only because it produces a convenient waveform.

Is adapter input inrush the same as device startup current?

No. Teams often use “inrush” for two different events, and the wrong definition can send a supplier in the wrong direction.

Adapter input inrush is measured at the AC input when the adapter is connected or switched onto the mains. It mainly concerns the adapter’s input stage and the upstream AC path.

Device startup current is drawn from the adapter’s DC output by the powered equipment. Motors, radios, heaters, printers, battery chargers, processors, and downstream capacitors can demand a short burst above their normal load. That event concerns output-current capability, voltage dip, current limiting, transient response, startup timing, and protection recovery.

The two events can occur during the same product power-up, but they are measured at different points and judged against different limits. A replacement adapter with improved AC-input inrush control can still fail to start a capacitive or motor load. A model that starts the DC load correctly can still nuisance-trip an upstream relay or breaker when several adapters energize together.

Write both requirements when the product needs both:

• AC-input startup waveform and inrush acceptance conditions
• DC-output startup load profile, voltage tolerance, and recovery behavior

The site’s power adapter RFQ guide shows where these input and output requirements belong in the wider sourcing package.

What can inrush disrupt in the finished product?

The adapter may survive its own startup while another component in the power path does not behave as intended. The affected item can be a fuse, circuit breaker, switch, relay, connector, PDU, UPS, generator-backed source, or control board that energizes the adapter.

Common symptoms include:

• a fuse opening during normal turn-on;
• a branch breaker or PDU protection device tripping when several units start;
• relay or switch contacts experiencing more startup stress than their rating allows;
• a UPS or controlled AC source entering protection;
• a brief line-voltage sag that resets another device on the same source;
• an intermittent failure that appears only during a quick restart or at one mains condition.

Multi-unit systems deserve their own scenario. Twenty adapters behind one relay are not equivalent to one adapter tested twenty separate times. Their pulses may overlap, partially overlap, or spread out depending on control timing and the AC waveform. Source impedance can also change the waveform. A conservative calculation is useful for screening, but the final system decision may require a simultaneous-start test with the real distribution hardware.

The same caution applies to production fixtures and burn-in racks. If every outlet is switched at once, the fixture can fail even though each adapter passes an individual functional test. Sequenced power-on, separate branches, or a different switching architecture may solve the system problem without changing the adapter, but the chosen approach must be verified under the intended cycle.

How should an OEM measure inrush current?

Use a controlled test plan and qualified personnel. AC mains measurement can expose hazardous voltage and high fault energy. The current probe, differential voltage probe, oscilloscope or power analyzer, AC source, fixture, and protective setup must all be correctly rated for the circuit and the expected pulse.

A repeatable plan should:

• Identify the exact adapter, hardware revision, cable, load, and sample.
• Define the input-voltage and frequency matrix required by the target markets.
• Define cold-start preparation and every relevant hot-restart off-time.
• Exercise enough switch-on points to find the relevant worst case rather than relying on one convenient capture.
• Use the actual device load where an electronic load cannot reproduce startup behavior.
• Capture input voltage and current together so the switching condition is visible.
• Record peak, duration, waveform, trigger method, probe settings, and test uncertainty.
• Repeat the test across enough samples to support the project’s risk and acceptance decision.

Instrument range matters. A probe chosen for a very high peak can lose useful resolution at lower current. A range that is too low can clip the event and report a false peak. Current probes may also require degaussing and zeroing, and their pulse-current rating, bandwidth, and rise time must suit the waveform. Tektronix’s current-measurement guidance explains these instrument tradeoffs.

Do not save only a screenshot with a cursor value. Preserve the setup and raw or exportable waveform when possible. That record lets another laboratory reproduce the result and helps the team understand whether a later difference came from the adapter, the source, the fixture, or the measurement method. The broader power adapter qualification test plan can bind this procedure to sample control, failures, retests, and release status.

How do you coordinate fuses, switches, relays, and breakers?

Compare the startup waveform with the exact protective and switching components in the product. A 40 A peak does not automatically require a fuse rated above 40 A because a fuse responds to current over time, not to the printed peak alone. The opposite mistake is also possible: a fuse with a suitable continuous-current rating may not tolerate a repeated startup pulse under the real ambient condition.

For a fuse, review its time-current characteristic, pre-arcing pulse capability or I²t data where applicable, voltage rating, breaking capacity, temperature derating, approval, and repetition behavior. SCHURTER’s fuse-link guide explains why time-current curves and I²t are part of pulse selection. The selected fuse and holder still need end-product verification under normal and fault conditions.

For switches, relays, and contactors, use the manufacturer’s rating for the relevant load and pulse, not only the steady current printed on the housing. Contact material, closing behavior, repetition rate, and expected service life can matter. For breakers and electronic PDUs, compare the relevant trip curve or protection behavior with both individual and aggregate startup events.

Do not increase a fuse or breaker rating merely to stop nuisance trips. That change can weaken protection of wiring or downstream components. If coordination fails, the engineering choices may include a different approved protective device, controlled sequencing, a suitable switching component, a different adapter design, or an active input-limiting approach. The correct fix depends on the complete safety and fault analysis.

What should the RFQ and acceptance record contain?

The RFQ should describe the system problem rather than ask only for “low inrush.” Give the supplier enough information to test the same event your product will encounter.

Include:

• target countries, input-voltage and frequency range;
• adapter format, power rating, exact output requirement, and expected load startup behavior;
• cold-start and hot-restart scenarios, including minimum off-time;
• upstream fuse, breaker, relay, switch, UPS, PDU, or AC-source details;
• number of adapters that may energize together and any sequencing;
• maximum permitted peak, duration, pulse energy, voltage sag, or nuisance-operation criterion when the system design has established one;
• required test matrix, sample quantity, waveform format, and report fields;
• exact-model and revision binding for qualification and production changes.

The acceptance record should connect the result to the purchase specification. Match the model and suffix across the datasheet, label artwork, drawing, sample, test report, and purchase order. Record any family rules or component alternatives that could change the input stage. If the supplier changes the limiter, rectifier, bulk capacitor, EMI filter, PFC stage, relay, firmware-controlled startup, or relevant PCB layout, require review under the project’s change-control rules.

A lower peak is not automatically the better adapter. It may involve a tradeoff with efficiency, startup time, thermal behavior, size, cost, or restart performance. Compare qualified models against the complete product requirement and ask for the conditioned evidence behind each value.

For a model-specific review, send Merryking the intended input markets, adapter wattage, load profile, upstream protection, simultaneous-start count, restart sequence, and required test conditions. The engineering response should identify whether an existing platform fits the requirement or whether sample testing is needed before production approval.