How to read a power adapter spec sheet

The efficiency number on a power adapter spec sheet is usually not a measurement. For an external AC-DC supply, the US Department of Energy sets a minimum average efficiency as a formula keyed to nameplate output power — so a 24 W adapter that claims exactly 86.2% and a 30 W adapter that claims exactly 86.95% are both quoting the DoE Level VI minimum for their wattage, copied onto the sheet. Once you know that, the rest of the document becomes much easier to read: some lines are obligations, some are design choices, and only a few are claims worth arguing about.

This page walks an external power supply spec sheet block by block, in the order the blocks appear. The values shown as typical are the ones we write on our own sheets as a power adapter manufacturer building mains-input units in the 18–36 W range; the regulatory values are cited. Read it against whatever sheet a supplier has sent you — the questions at the end are the ones worth putting in an email.

Efficiency: a formula, not a claim

DoE Level VI has applied to external power supplies sold in the US since February 2016. For a single-voltage external AC-DC supply at basic voltage — meaning nameplate output of 6 V or more — the minimum average active-mode efficiency is:

Efficiency here means the average of four readings — at 25%, 50%, 75%, and 100% of rated load — not a single best-case point. Worked out for common wattages:

If the sheet in front of you states a number that matches this table to the decimal, you are looking at the legal floor written down, not bench data. That is not dishonest — it is how the industry writes these documents — but it means the sheet tells you the product is compliant, not how good it is.

If efficiency actually matters to your thermal budget or your energy claims, ask the supplier for the test report with the four load points on it. A manufacturer has one; a trading company has to go and request it, and the delay tells you which you are dealing with.

Where Level VI does not reach

Two common cases fall outside the standard, and buyers routinely compare across them without noticing:

• Car chargers and other DC-input supplies. The DoE definition covers circuits converting household mains current; a DC-DC-only supply that cannot run on household mains is excluded. A 12/24 V car charger therefore has no Level VI obligation, which is why one may honestly specify 70% minimum efficiency next to a wall adapter’s 86%. The parts are not comparable on that line.
• AC-AC linear transformers. These are external power supplies by definition, but a linear design often cannot meet the Level VI curve. A sheet that quotes a healthy-looking efficiency and then states it does not comply with Level VI is flagging a real problem you must resolve before selling the product into the US — not a footnote.

Input block

Output block

A ripple figure quoted without a bandwidth limit is not comparable to one that states it. Two suppliers can differ by a factor of two on the same hardware purely through measurement setup, so read the note under the table before you read the number in it.

Protection block

Protection is the block most often written as prose and least often written with numbers. Presence is not a specification — trip points are. What a complete sheet names:

Two questions to ask about any protection line. Where does it trip, and does it latch or auto-recover? A latching supply that needs a mains cycle to reset is a field-service call in an unattended installation; an auto-recovering one may oscillate into a persistent fault. Neither is wrong, but the sheet should say which you are getting.

Safety block

These three lines are the ones a safety agency will look at, and the ones a serious factory can produce test records for on request.

Hi-pot is a 100% production test, not a type test — every unit should see it on the line, and the sheet is describing what the line is set to.

Ask whether the test is logged per unit and whether you can have the records. The answer separates a factory from a trading company faster than most questions in a supplier audit. How we run ours is described on our quality page, and what we build is on the manufacturing page.

EMC and immunity

The EMI line is usually one entry — FCC Part 15 Class B for a US consumer-adjacent product. The immunity block is where sheets get thin: a standards list with no levels and no performance criteria tells you nothing. A complete block looks like this:

Performance criterion A means the unit must keep working normally during the disturbance — not merely survive it and recover. A sheet naming the standard but not the criterion is leaving out the part that matters.

Mechanical and environmental

The mechanical block is short and is where most late-stage surprises live, because it is the block buyers skim.

• Output cable — e.g. 2468 20AWG, 1800 mm, black. The wire type and gauge set the voltage drop at the far end. On a long lead at full current, an undersized conductor can put the load below its minimum input voltage while the adapter itself is perfectly in spec.
• Connector — barrel size and polarity, or to your drawing. Confirm polarity in writing; centre-positive is common but not universal. On an OEM or custom build this line is yours to set, so set it explicitly rather than inheriting a default.
• Dimensions and weight — e.g. 76 × 51 × 31 mm. Check blade-to-body clearance against the outlet strips your customer actually uses.
• Operating range — typically 0–40 °C, 10–90% RH, with storage at −20–60 °C. If your product lives in an unheated cabinet or a vehicle, 0 °C is a real constraint.
• Acoustic noise — under 25 dBA at 1 m. Relevant in ward, studio, and office installations, where a faint transformer whine generates complaints no electrical test would predict.

What should make you suspicious

A spec sheet is a manufacturing document. When it reads like a brochure, it usually was not copied from a test report:

1. Efficiency at one load point instead of the four-point average the standard requires.
2. Ripple with no bandwidth limit or measurement setup stated.
3. No no-load figure — an odd omission, since it is a mandatory limit.
4. Protection listed without trip points, or with no statement of latch versus auto-recovery.
5. An EMC section with standard numbers but no levels and no performance criteria.
6. No revision number, date, or change history. A sheet that has never been revised has usually never been tested against.
7. Certification logos without file numbers. A UL mark is checkable; ask for the file number and check it.

None of these prove a bad part. All of them tell you the document was assembled to close a sale rather than to control a build — which means the thing you are agreeing to buy has not actually been written down yet.

Using this to evaluate a supplier

The sheet is a proxy for how a supplier works. A contract manufacturer that writes its own sheets can answer follow-up questions in the same call: where OCP trips, whether hi-pot is logged per unit, which load points the efficiency average came from. A supplier reselling someone else’s product has to go and ask, and the answers come back in fragments.

Three requests separate the two quickly, and none of them are unreasonable to make before a first order:

1. The four-point efficiency test report, not the sheet’s summary figure.
2. The hi-pot log format — per unit, with the trip setting on it.
3. A sheet revision history with dates, showing the document has been changed at least once.

We build wall-mount, desktop, and USB-C adapters on our own lines, and write the sheets that go with them. If you are comparing our sheet against an incumbent supplier’s, the comparison we would rather you make is on these three documents than on the summary numbers.