Charger Surge Immunity: Why a Batch Survives One Market and Fails in Another

A batch that survives one market and dies in another

Same charger, same factory, same test report. In one country it runs three years without a complaint. In another, the distributor reports returns climbing month by month, and the units that come back are not dead on arrival — they worked for half a year and then stopped. When that pattern shows up, the cause is rarely the power stage. It is what the grid did to the power stage.

Lightning, switching transients from air conditioners and elevators, and the surge that follows a blackout all arrive at the charger’s input as a short, high-energy voltage spike. A charger that has never been asked to survive one will usually fail in a way that leaves no obvious burnt mark — a cracked varistor, a shorted bridge diode, a controller that latches up. The buyer sees “quality problem”. The factory sees nothing wrong with its production data. Both are right, and the gap between them is a missing line in the specification.

What IEC 61000-4-5 actually specifies

Surge immunity is tested to IEC 61000-4-5. The standard defines a 1.2/50µs voltage waveform (and an 8/20µs current waveform) delivered through a combination wave generator, applied between line and neutral, and between line/neutral and earth. Tests are graded by severity:

• Level 1 — 0.5kV, for well-protected environments.
• Level 2 — 1kV, typical indoor installation in a reasonably stable grid.
• Level 3 — 2kV, the level most commonly written into charger specifications.
• Level 4 — 4kV, harsh environments: overhead distribution, frequent storms, industrial neighbours.
• Class X — a custom level agreed between buyer and supplier; it exists precisely because “3kV” is a legitimate requirement in some markets.

Two things about that list catch people out. First, the level applies to both coupling modes, and line-to-earth is the harder one for a two-wire charger — there is no earth pin, so the spike has to be absorbed by the isolation barrier. Second, the standard does not decide what counts as a pass. That is a separate document, and it is where most certificate surprises come from.

Performance criteria: the part that gets skipped

IEC 61000-4-5 references the performance criteria in IEC 61000-4-1, and the four categories matter more to a buyer than the voltage level:

• Criterion A — normal performance during and after the test. No deviation.
• Criterion B — temporary degradation that clears itself; the unit recovers without intervention.
• Criterion C — degradation requiring operator intervention, such as a power cycle.
• Criterion D — damage that is not recoverable.

Someone quoting “passes 2kV surge” has told you half the story. A charger that drops its output for 200ms and recovers is Criterion B and counts as a pass. A charger that needs to be unplugged and plugged back in is Criterion C — also technically a “pass” under many test plans, and completely unacceptable in a bedside charger. Ask for the criterion, not just the voltage.

Don’t confuse surge with EFT or ESD

Three tests get conflated in datasheets and they fail for different reasons:

A charger can pass 4kV surge and still fail 1kV EFT, because the two inject energy differently — the burst test is a train of fast pulses that finds poor layout and inadequate filtering, where the surge test finds insufficient clamping energy. If your returns come from a market with heavy motor load rather than storms, EFT is the test worth raising.

Why small chargers are structurally exposed

Surge protection costs board area and money, and both are scarce in a 30W wall charger. The protective parts are a metal-oxide varistor across the input, sometimes a gas discharge tube on higher-end designs, a fuse that must survive the surge without opening, and adequate creepage across the isolation barrier. In a compact GaN design the varistor is often a small disc, and the fuse is a slow-blow type rated just above the working current.

The consequence: a charger can be perfectly safe and perfectly certified, and still be undersized for a market where the grid delivers repeated 2kV events. Certification proves the design met the level declared in the report — it does not prove the level suits your market.

What to write into the purchase specification

• Test level and coupling: “IEC 61000-4-5, Level 3 (2kV line-to-line and line-to-earth)” — name both modes explicitly.
• Performance criterion: “Criterion B or better at the specified level; no latch-up, no fuse opening, no output interruption longer than 100ms.”
• Repeated-strike survival: ask for the number of strikes applied. A single strike at 2kV is a different product from ten strikes at 2kV with the unit still working.
• Surge current rating of the varistor: the datasheet number (often 8/20µs, in amps) tells you the real energy budget.
• Field-return data by market: if the factory has shipped into the market before, its return rate per region is more informative than any test report.

Reading the EMC report for this specific item

The surge result lives inside the EMC test report, usually as a table row under “Immunity”. Three checks take two minutes:

• Does the report list IEC 61000-4-5 at all? Plenty of low-cost reports cover only emissions (EN 55032) and skip immunity entirely — that is a scope problem, not a failure.
• Does the row state the test level and the criterion, or only “Pass”? “Pass” with no level is unverifiable.
• Does the model number in the report match the model on the purchase order, including the plug variant? A surge result for the EU plug does not transfer to the UK fuse version.

Where chargers and surge protection meet in the field

An honest note on expectations: no wall charger can replace a proper surge protective device at the distribution board. What a charger can do is survive the events that get past the building protection — the residual spikes that appear at the socket after a nearby strike, or the transient from a compressor starting on the same circuit. Raising the specification from 1kV to 2kV, and insisting on a defined criterion, changes the return rate in storm-prone markets. Raising it to 4kV changes the cost and the size. Between those two, the right answer depends on which market you are selling into, not on which number looks better on a box.

One more practical point for markets with unstable supply: a charger that survives surges is not the same as a charger that behaves well during brownouts. Under-voltage behaviour — whether the unit restarts cleanly when mains returns, or sits latched — is worth asking about separately, and it is usually visible the moment you test a sample on a variable transformer.

Specifying surge immunity for a market

Two inputs decide the answer: the destination market’s grid profile, and how the product will be installed (fixed socket, extension lead, behind furniture). Send both, plus the output rating and the certification your retail channel requires, and the reply should state the surge level, the criterion, the varistor rating, and which plug variants the test report covers. If a quotation says “high surge protection” without a level, it has not specified anything yet.

The product range this applies to covers GaN and PD chargers from 20W to 100W, in Type-C, USB-A and multi-port formats, with market-specific certification packs — see the GaN 100W charger for a representative multi-port layout, or the certification checklist for the documents worth collecting alongside the EMC report.