Back to Home
earpieces

Your Earpiece Spends Twenty Seven Milliwatts. The Speaker Gets Half a Nanowatt.

An AirPods Pro 3 bud holds 0.221 watt hours and runs eight hours, which means the whole device averages 27.6 milliwatts. The driver's share of that, at ordinary listening levels, is under a thousandth. Here is where the rest goes: amplifier quiescent current, four always-on microphones, a 96 kHz cancellation loop, a case that loses energy twice, and a sleep state most vendors will not publish.

Disclosure: Some links in this article are affiliate links. If you purchase through them, we may earn a small commission at no extra cost to you. This helps support our independent editorial coverage.

Twenty Seven Milliwatts, and Almost None of It Reaches Your Ear

Open an AirPods Pro 3 far enough to read the cell and the vagueness evaporates. iFixit measured 0.221 watt hours in each bud. Apple advertises eight hours of listening per charge. Divide one figure by the other and the entire device averages 27.6 milliwatts across the session. Every filter, every microphone, every radio packet and every amplifier inside that sealed plastic shell lives inside twenty seven milliwatts.

Now compare that against the work the driver is asked to do. A 70 dB SPL tone at the eardrum amounts to roughly half a nanowatt of radiated acoustic power, spread across the half square centimeter the eardrum presents. Half a nanowatt against 27.6 milliwatts is one part in fifty million. Everything else is overhead, and the overhead is the product.

What the Driver Actually Consumes

Take a typical earbud driver rated 105 dB SPL at one milliwatt. Achieving 70 dB from it requires 0.32 microwatts of electrical drive. A loud passage peaking at 100 dB needs 0.32 milliwatts. Music averaging 30 microwatts of electrical input while peaking ten times higher is ordinary listening, which means the transducer draws something between one ten thousandth and one thousandth of the system budget.

That ratio is the story of earbud engineering. You are not designing an audio amplifier. You are designing a computer that happens to output audio, and the audio is nearly free.

The Amplifier Is Worst Exactly Where It Lives

Class D efficiency curves published in datasheets usually start at a watt and climb. Down at 30 microwatts of output, a switching stage is governed almost entirely by its own overhead: gate charge, switching transitions, modulator quiescent current, comparator bias. Efficiency at that operating point can fall under ten percent, and the number hardly matters because the loss is a fixed cost that does not shrink with the signal.

An off the shelf three watt class D part idles at 2.4 milliamps from five volts. That is twelve milliwatts before it plays a note, and nearly half our total budget. Earbud silicon is far better behaved, often idling in the low hundreds of microamps, but the shape of the problem holds: fixed draw dominates variable draw by orders of magnitude. Some designs gate the output stage between pulse trains to cut it. The audible cost is a soft floor of idle noise that measures well and occasionally does not.

Supply topology adds another line item. A single lithium cell runs from 4.2 volts down to 3.2, and a driver asked for 100 dB plus headroom wants more swing than that at the end of the day. A charge pump or boost stage supplies the extra rail, and boost converters are efficient at their design point and mediocre at 200 microamps. Headroom costs conversion, conversion costs a few hundred microamps, and the sum arrives in the battery report as an hour of listening you never had.

The ANC Tax Is Measurable and Larger Than You Think

Sony publishes 8 hours with noise cancelling on and 12 hours with it off for the WF-1000XM5. If those figures were taken at matched conditions, the arithmetic is unforgiving: total power with ANC is 1.5 times total power without, so the cancelling loop consumes one third of the entire budget. Roughly 40 milliwatts in place of 27, or a 33 percent tax for the feature you probably bought the earbud for.

Where does it go? Hybrid ANC runs at least four microphones, two feedforward and two feedback, each digitized continuously. The feedback path demands a short acoustic delay, which pushes the loop to 96 kHz or higher, so anti alias filters, decimators and the loop filter all run faster than the music does. Adaptive filters tracking the leak between tip and canal update thousands of times a second. Then there is the cancellation signal itself, a second amplifier channel dumping anti phase energy into the same driver trying to reproduce music, and that amplifier pays the same fixed overhead as the first.

Turn cancelling off and the microphones often stay powered for transparency, which is why the gap between ANC on and ANC off is smaller than the gap between transparency on and everything off. The switch on your phone controls the algorithm. It does not always control the hardware.

Microphones Never Sleep

A modern PDM MEMS microphone draws somewhere between 100 and 400 microamps at 1.8 volts, so call it a third of a milliwatt each. Four of them cost 1.3 milliwatts, five percent of a tight budget, and that is before beamforming or wind noise processing touches a sample. Add a bone conduction accelerometer for voice pickup and another few hundred microamps join the pile. Add always listening wake word detection and the number climbs again, though good implementations keep that function in the tens of microwatts by running a tiny always on classifier and waking the main core only on a candidate.

Here a spec sheet lies by omission. Two earbuds with identical drivers and codecs can differ by an hour of runtime because one vendor powers down its transparency path when nothing plays, and the other leaves four microphones listening to an empty room.

The Radio Is a Duty Cycle Problem

Bluetooth Classic carries audio in 625 microsecond slots. A 2-DH5 packet holds 681 bytes of payload, which at a typical bitpool carries three or four SBC frames. Between packets the transmitter is off, and off is nearly free. Average transmit power for high quality A2DP lands in the low single digit milliwatts, well below the amplifier's idle draw and well below ANC.

LE Audio shifts the accounting rather than reducing it. LC3 frames at 7.5 or 10 milliseconds mean more wakeups per second than a fat classic packet, and each wakeup costs a fixed energy toll in synthesizer settling, PLL lock and receive window overhead. Shorter frames buy latency and robustness. They do not buy battery, and designers who assume the leaner codec must be the cheaper system get surprised by the current probe.

The Case Is Where the Real Losses Hide

Charging one lithium cell from another is not a wire. A contact or inductive path with a converter in the middle typically returns 80 to 90 percent, and the earbud's own charger chip takes a few percent more. Stack those against the advertised totals and something interesting appears. A 1.334 watt hour case supplying three full charges of 0.221 watt hours delivers 1.326 watt hours, which leaves no room whatsoever for conversion loss.

Some of that is rating practice, some is capacity quoted at a gentler discharge rate than real use, and some is simply that the case total assumes conditions nobody achieves. The practical consequence matters more than the accounting. The case figure is the least trustworthy number on the box, and it is the one shoppers compare.

Standby Is the Quiet Killer

Twenty seven milliwatts for eight hours is respectable. Four milliwatts of idle drain is catastrophic, because idle happens for days. At that rate the 0.221 watt hour cell empties in about 55 hours, which explains why a wireless earbud found in a bag after a long weekend is flat even though it was switched off. The fix is a genuine deep sleep state measured in microamps, plus a case that tops the bud up whenever the lid closes. Every manufacturer claims this. The ones who mean it publish standby current. The ones who do not publish standby time in hours, which hides the same measurement behind a friendlier unit.

Where the Hours Actually Come From

Not from codec marketing. A high bitrate codec costs the radio more than a lean one, but the delta is a fraction of a milliwatt in most implementations. Real gains come from four places, in descending order of size: trimming ANC filter order or sample rate where the acoustics allow it, duty cycling microphones properly, choosing an amplifier whose quiescent current is specified at your actual output level rather than at one watt, and eliminating leakage in the state machine that decides when things wake up.

That last one is unglamorous and frequently botched. A stray 20 microamps is 0.07 milliwatts, trivial in isolation, enormous when the device spends 130 hours a week pretending to be off.

Three Things You Can Measure at Home

Charge both buds, disable ANC and transparency, fix the volume, and run a podcast until the first bud dies. Repeat the next day with ANC on. The ratio gives you your specific ANC tax, and it will shift with content because cancellation difficulty is not constant.

Then test the case. Note the percentage, drop the buds in for exactly ten minutes with the lid shut, and read the increase. A case delivering noticeably less than its watt hours suggest is losing energy in conversion, which is normal, or in standby, which is not.

Finally, leave a bud out of the case, disconnected from everything, for a day. Whatever it loses overnight is the floor of the design. Twenty seven milliwatts while working is an achievement. One milliwatt while waiting is good engineering. Nothing at all while waiting is a marketing claim, and now you can tell the difference.

An Earpiece Is a Power Budget With a Speaker Attached

The next time a review tells you a pair sounds better, ask what changed in the current draw. Drivers moved maybe two tenths of a milliwatt. The rest of the difference lives in filters, microphones, radio scheduling and sleep states, all hidden behind a battery bar that reports whole hours. The acoustics were always the easy part. The twenty seven milliwatts are the design.

#earpieces#power budget#ANC battery drain#MEMS microphone#class D efficiency#quiescent current#LE Audio power#AirPods Pro 3 battery#WF-1000XM5 battery#audio physics

Related Articles

blutoof.com

Your premier source for Bluetooth technology news, reviews, and guides. From premium headphones to hands-free earpieces — we cover it all.

Affiliate Disclosure: Blutoof.com is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Amazon.com. When you purchase through links on our site, we may earn an affiliate commission at no extra cost to you.

© 2026 Blutoof.com. All rights reserved.

Bluetooth® is a trademark of Bluetooth SIG, Inc.