Efficiency and heat
Why a charger gets warm, what's normal, and how much a few percent really matters
A charger warms up because no conversion is perfect: the power it doesn't deliver to your iPhone becomes heat. A good 20W charger converts roughly 85–90% of what it draws from the wall, so at full output it sheds about 2–3.5 watts as heat. Independent tests measured Apple's 20W adapter at 42–49.5°C on its surface after an hour at full load in a 25°C room. Warm to the touch is normal; a charger too hot to hold, discoloured or smelling of burning is not — unplug it.
In 30 seconds
- Efficiency = power out ÷ power in. The rest is heat.
- US and EU rules require about 85.5% average efficiency for a 20W adapter and no more than 0.10 W drawn with nothing connected. 12
- Apple's 20W adapter measured 85.5–89.7% efficient at 220 V and 0.021 W at no load in ChargerLAB's 2023 test. 3
- A warm charger is normal. The same test found 42.1°C on the front and 49.5°C on the back after an hour at 20W in a 25°C room. 3
- Small temperature differences prove little about lifespan. ERONIX will publish Fermi20's efficiency and temperature with full method, not as a longevity promise.
In three minutes
The arithmetic
Efficiency, written η (eta), is:
η = P_out ÷ P_in × 100%
Everything not delivered becomes heat inside the charger:
P_loss = P_in − P_out = P_out × (1/η − 1)
A worked example at 20W output:
| Efficiency | Power drawn from the wall | Heat inside the charger |
|---|---|---|
| 85% | 23.53 W | 3.53 W |
| 88% | 22.73 W | 2.73 W |
| 90% | 22.22 W | 2.22 W |
Going from 85% to 90% removes about 1.3 W of heat — roughly a third less, which a hand can feel on a small enclosure. In electricity it is trivial: two hours a day at full load for a year is about 1 kWh. Efficiency matters at 20W mainly for comfort, temperature and component stress, not running cost.
What the rules require
Both the United States (Department of Energy Level VI, since 10 February 2016) and the European Union (Regulation 2019/1782, since 1 April 2020) use the same formula for single-voltage external power supplies between 1 and 49 W. 12
Bar chart: 85.5 percent basic-voltage minimum, 83.1 percent low-voltage minimum, 81.4 percent at 5 volts 3 amps.
| Rule | Basis | Minimum at 20 W |
|---|---|---|
| US DoE Level VI / EU 2019/1782 — basic voltage (applies at the 9 V, 20 W nameplate) | 0.071·ln(P) − 0.0014·P + 0.67 | 85.47% |
| Same — low-voltage formula (outputs below 6 V at ≥ 550 mA) | 0.0834·ln(P) − 0.0014·P + 0.609 | 83.08% |
| Same — low-voltage at the 5 V, 3 A (15 W) point | — | 81.39% |
| No-load power, both regions, ≤ 49 W | — | ≤ 0.10 W |
| EU 2025/2052 (applies from 14 Dec 2028), basic voltage, average | draft annex | 85.97% |
| EU 2025/2052, minimum at 10% load, basic, non-adaptive | draft annex | 79.97% |
| EU 2025/2052, no-load | draft annex | ≤ 0.075 W |
The 2025/2052 values come from the Council's pre-adoption annex and should be confirmed against the final Official Journal text. 456 Regulation 2019/1782 sets no minimum at 10% load; it only requires that figure to be declared. 2 We found no Indian minimum efficiency rule for phone chargers in this research.
Apple's 20W adapter as a reference
ChargerLAB's measurements of one Apple 20W A2940 unit 3:
| Measure | 220 V 50 Hz | 110 V 60 Hz |
|---|---|---|
| Efficiency (range across loads) | 85.52–89.73% | 84.65–88.73% |
| No-load draw | 0.021 W | 0.018 W |
| Surface temperature after 60 min at 20 W, 25°C room | front 42.1°C, back 49.5°C | front 46.0°C, back 52.9°C |
ChargerLAB described the efficiency as "average" and the temperatures as within IEC 62368 limits. Techtest.org measured an earlier EU unit (A2347) at 83–88.8%. 7 Note the pattern: chargers run slightly more efficiently and cooler on higher mains voltage, which favours India's 230 V supply. That is why ERONIX tests at 230 V 50 Hz first, then at low-line voltages.
Why it gets warm — and what's normal
At full output, a 20W charger dissipates two to three-and-a-half watts in a sealed box the size of a matchbox. Heat flows to the surface and into the socket. Readings climb for the first 30–60 minutes and then level off (thermal equilibrium). What you feel depends on:
- Load. An iPhone draws close to 20W only in the first part of a charge; as it fills, the charger cools. 3
- Room temperature. A charger in a 35°C room in May runs hotter than in a 25°C lab.
- Mains voltage. Lower input voltage, slightly more heat. 3
- Where it sits. Behind a sofa, under a pillow or in a crowded multi-plug, heat can't escape.
- Where you measure. The hottest spot (often the side nearest the transformer or the back face) can be 5–10°C above the front. 3
Safety standards for audio/video and IT equipment (IEC 62368-1, adopted in India as IS/IEC 62368-1:2023) set limits on how hot accessible surfaces may become, depending on material and how long a person is expected to touch them. 89 We have not verified the specific numeric limits against the standard's text for this page, so we do not quote them.
Does a warmer charger shorten its life — or my battery's?
Electrolytic capacitors and other components do age faster at higher temperatures, which is why good designs derate parts and manage heat. But a few degrees' difference between two well-built chargers, measured once, does not by itself prove one will outlast the other; that would need long-duration reliability testing on many units. ERONIX does not make lifespan claims from case temperature.
The charger's temperature is also separate from your iPhone's. iPhone manages its own battery temperature and may limit charging above 80% when warm. 10 Charge in a cool place, keep the phone out of direct sun, and remove thick cases if it gets hot during charging.
For engineers
E1. Average active efficiency
Average active efficiency is the arithmetic mean of efficiency at 25%, 50%, 75% and 100% of nameplate output current. For a multi-voltage USB PD adapter, the test points and which formula applies at each output voltage depend on the applicable test procedure (DoE Appendix Z; the EU harmonised method). The 9 V/2.22 A nameplate falls under basic voltage; the 5 V/3 A output meets the EU low-voltage definition (< 6 V, ≥ 550 mA). 2 Test at both. The exact procedure clauses were not verified in this research.
E2. ERONIX efficiency and no-load method
| Parameter | Setting |
|---|---|
| AC source | Programmable, low-distortion; 230 V 50 Hz primary; 100 V 50 Hz, 115 V 60 Hz, 264 V 50 Hz secondary |
| Power meter | Calibrated, resolution ≤ 1 mW at no load; integrating mode for no-load (≥ 5 min after 30 min warm-up) |
| Load | Electronic load at the cable end (reference cable resistance recorded) and at the receptacle (cable loss reported separately) |
| Points | 10, 25, 50, 75, 100% of each PDO's maximum current; average active efficiency computed |
| Thermal | Warm-up 30 min at 100% before efficiency sweep; ambient 25 ± 2°C |
| Samples | 3 production units; mean, min, max |
| Uncertainty | Combined meter + load uncertainty stated per point |
E3. Thermal method
| Parameter | Setting |
|---|---|
| Ambient | 25 ± 1°C still air (draught-free enclosure); repeat at 35°C for Indian summer conditions |
| Mounting | Plugged into a standard Indian wall socket on a vertical board, as in use |
| Load | 9 V, 2.22 A continuous for 120 min; record every 10 s |
| Sensors | Thermocouples (type T or K, fixed with thermally conductive adhesive) on front, back, sides and plug face; thermal camera for hotspot location with emissivity set for the enclosure finish |
| Reported | Hotspot, front, back, rise above ambient, time to equilibrium; camera images with scale and emissivity |
| Not done | Decorative false-colour images without scale |
Myths
"A hot charger is faulty." Warm is normal under load. Too hot to hold, deformed, discoloured or smelling of burning is not.
"A more efficient charger charges faster." Speed depends on the power the iPhone requests, not the charger's efficiency.
"GaN chargers always run cooler." Material can help; design and measured result decide. At 20W, Apple's silicon adapter is efficient. 11
"Leaving a charger plugged in wastes a lot of power." A good one draws a few hundredths of a watt with nothing attached — Apple's measured 0.021 W, about 0.18 kWh a year. 3
Questions people ask
How hot is too hot for a phone charger? Independent tests show a well-made 20W charger reaching about 42–53°C on its surface after an hour at full load, depending on input voltage and face measured. Safety standards set limits for touchable surfaces; if a charger is too hot to keep your hand on comfortably or shows any damage, unplug it. 3
Does a warm charger hurt my iPhone's battery? The charger's heat stays in the charger. Your battery is affected by the iPhone's own temperature; iPhone manages this and may pause charging above 80% when warm. 10
Why is my charger warmer in summer? Higher room temperature leaves less margin for heat to escape, so surface temperature rises by roughly the same amount as the room.
Is 230 V better than 110 V for a charger? Most chargers are slightly more efficient and cooler at higher input voltage; ChargerLAB's Apple 20W results show this. 3
What efficiency will Fermi20 have? Measurement in progress.