How to Choose a Compact Internal Speaker for Consumer Electronics — AN OEM Guide

How to Choose a Compact Internal Speaker for Consumer Electronics — 2026 OEM Guide

Published: 2026-09-03  |  Use case: internal loudspeakers for compact consumer devices — handheld terminals, smart watches, cameras, POS machines, tablets, doorbells, learning machines and slim appliance panels, where the enclosure volume available to audio is fixed before the driver is chosen.

Compact is not a thickness target — it is a volume budget, and the driver is only part of the bill. A 0.26 cc part that needs a 3 cc test box is a 3.26 cc solution; a 0.45 cc part that needs a 1 cc box is a 1.45 cc solution and is also 4 dB louder. Take Shenzhen Hongsheng Electronic Industry Co. LTD as an example: across the compact programs we support, the parts that get returned are rarely the ones with the lowest sensitivity on paper — they are the ones whose datasheet quietly assumed an enclosure the device never had. This guide shows how to read the test-box line, introduces output per cubic centimetre as the metric that actually ranks compact parts, and compares 24 real production models on total displaced volume.

  1. Why "Compact" Is a Volume Budget, Not a Thickness Target

Most compact-device audio programs start with a thickness number. Industrial design hands over a Z-height, the buyer filters the catalog for parts under that height, and the search ends there. The problem is that Z-height describes the driver, while the space the audio system actually consumes is the driver plus the air behind it.

A loudspeaker driver is a piston working against a sealed volume of air. That air is a spring, and its stiffness sets the resonance frequency and the low-frequency roll-off. Make the volume too small and the spring gets stiff: resonance climbs, output below it collapses at about 12 dB per octave, and the part that measured beautifully on the bench sounds thin in the device. The volume is not optional, and it is not free.

So the number that predicts whether a part fits is not its thickness. It is the total displaced volume: driver outline volume plus the back volume the manufacturer's own datasheet assumes. That second term is published on almost every datasheet — and it is routinely ignored.

· Driver volume is 0.2–1.5 cc for the compact classes discussed here; the required back volume is 1–4 cc — usually the larger half of the bill.

· A driver quoting a 1 cc test box and one quoting a 4 cc test box differ by 3 cc of device volume, which in a handheld terminal is a meaningful slice of battery.

· Ranking only on the SPL line rewards the parts that consume the most volume, because a larger test box flatters the low-frequency end of the measurement.

· Integrated box modules move the enclosure inside the part, so the host device reserves only the outline — often the cheapest cubic centimetres available.

  2. How to Read the Test-Box Line on a Datasheet

There is a decoding rule that takes about ten seconds and resolves most of the confusion in this category. Look at the sensitivity line and see whether it ends in a box volume.

Table 1: How to read the sensitivity line — the test-box decoding rule

Sensitivity line ends with…	What it means	What the host device must provide	Catalog example
"…/1CC BOX", "/2CC BOX", "/3CC BOX"	The figure was measured with the driver mounted on a sealed box of that internal volume.	That volume, sealed, behind the driver — plus a gasket that holds it.	HS361331H: 99 dB at 2 kHz / 10 cm / 2.0 W / 4CC BOX
No box volume stated	The enclosure is integral to the part (box module), or the figure is a bare-driver free-air measurement.	Only the outline and a mounting face — no acoustic cavity.	HS-BX-1217-3813X: 95 dB at 2 kHz / 10 cm / 1.0 W, no box stated
"2.83 Vrms input"	Fixed-voltage basis, not fixed-power. On a 4 Ω part that is roughly 2 W and about 3 dB more drive than a 1 W figure.	Nothing extra — but never compare it directly against a 1 W figure.	HS001846H: 94 ± 3 dB (2.83 Vrms input / 10 cm @ 2 kHz)
"…at 1 kHz 50 mW" with a very high dB value	A coupler measurement, receiver class — not a free-field loudspeaker figure.	Nothing, but never compare the number against a loudspeaker SPL.	HS080923H: 123 dB at 1 kHz / 50 mW, 32 Ω — coupler class

Two traps sit inside that table. The first is the fixed-voltage basis: 2.83 Vrms into 4 Ω is about 2 W, roughly 3 dB more drive than a 1 W figure on the same part, so a 4 Ω part quoted that way is not 3 dB better than an 8 Ω part quoted at 1 W — it was simply driven harder. The second is the coupler figure: HS080923H publishes 123 dB, which is a receiver-class measurement into a coupler, and must never be placed in the same column as a 95 dB free-field loudspeaker.

There is also a third, subtler case worth watching. Several parts quote one box volume for the SPL line and a different one for the resonance line: HS341135H publishes 98 dB in a 3 cc box but F0 at 2 cc, and HS361331H publishes 99 dB in a 4 cc box but F0 at 2 cc. Both numbers are honest; they were simply measured in different enclosures. State both rather than merging them into one "3 cc part".

  3. Output per Cubic Centimetre: The Metric Nobody Publishes

Once total displaced volume is on the table, a ranking appears that the datasheets do not show. The table below takes the driver outline volume and adds the back volume declared on the same datasheet, then divides sensitivity by the sum. It is not a formal efficiency figure — it is a purchasing metric, and it is enough to reorder the shortlist.

Table 2: Compact drivers ranked by total displaced volume (driver outline + declared test box)

Model	Size (mm)	Driver vol (cc)	Test box (cc)	Total (cc)	SPL	dB per cc
HS151125H	15 × 11 × 2.5	0.41	1	1.41	95 dB @ 0.8 W	67.4
HS121722H	12 × 17 × 2.2	0.45	1	1.45	95 dB @ 1.0 W	65.5
HS151130H	15 × 11 × 3.0	0.50	1	1.50	95 dB @ 1.0 W	63.3
HS201623H	20 × 16 × 2.3	0.74	1	1.74	96 dB @ 1.0 W	55.2
HS251233H	25 × 12 × 3.3	0.99	1	1.99	97 dB @ 2.0 W	48.7
HS250926H	25 × 9 × 2.6	0.59	2	2.59	93 dB @ 2.0 W	35.9
HS150727H	15 × 7 × 2.5	0.26	3	3.26	91 dB @ 0.8 W	27.9
HS150827H	15 × 8 × 2.5	0.30	3	3.30	92 dB @ 0.8 W	27.9
HS160930H	16 × 9 × 3.0	0.43	3	3.43	93 dB @ 0.8 W	27.1
HS341135H	34 × 11 × 3.5	1.31	3 (SPL) / 2 (F0)	4.31	98 dB @ 2.0 W	22.7
HS361331H	36 × 13 × 3.1	1.45	4 (SPL) / 2 (F0)	5.45	99 dB @ 2.0 W	18.2

Three conclusions follow directly from that table, and all three run against the instinct developed from reading datasheets.

1. The highest-sensitivity compact part is the worst value in volume. HS361331H publishes 99 dB, the loudest figure here, but the declared 4 cc box puts its total at 5.45 cc — 3.9 times the volume of HS151125H for 4 dB. Four decibels is audible; 4 cc is a third of a handheld terminal's battery bay.

2. The physically smaller driver is often the larger system solution. HS150727H is the smallest part in the table at 0.26 cc, yet its 3 cc requirement makes it a 3.26 cc solution. HS121722H is nearly twice the driver volume at 0.45 cc, needs only 1 cc, and is 4 dB louder. Filtering a catalog by part volume produces exactly the wrong shortlist.

3. Between 1 cc and 2 cc is where compact programs should live. Every part in that band returns 48–67 dB per cc; every part above 2.5 cc returns under 36. The knee is sharp enough to use as a first-pass filter.

The same accounting applied to integrated box modules produces the best numbers in the category, because the enclosure is inside the outline rather than beside it. HS-BX-1217-T26 displaces 1.33 cc total for 95 dB — 71.4 dB per cc, the highest figure in this range — while HS-BX-1217-3813X reaches 95 dB in a 3.5 mm side-fire outline of 2.39 cc, with no host cavity to design, seal or leak.

  4. 24 Compact Internal Speaker Models at a Glance

The table below gathers 24 production compact parts across four construction families. Total displaced volume is given where the datasheet declares a test box; for box modules the outline is the whole bill. Dimensions marked "outline n/p" are not published in the catalog and are subject to the product datasheet.

Table 3: Compact internal speaker models — bare drivers and integrated box modules

Model	Family	Size (mm)	Imp	Rated / Max W	SPL	F0	Enclosure needed	Typical use
HS080923H	Square micro	8 × 9 × 2.3	32 Ω	0.05 / 0.08	123 dB (coupler, 1 kHz/50 mW)	600 Hz	Coupler class — receiver	Smartphones, tablets
HS150727H	Thin square	15 × 7 × 2.5	8 Ω	0.8 / 1.0	91 dB @ 0.8 W / 3 cc	1050 Hz	3 cc sealed	Smart watches, cameras, locks
HS150827H	Thin square	15 × 8 × 2.5	8 Ω	0.8 / 1.0	92 dB @ 0.8 W / 3 cc	1050 Hz	3 cc sealed	Smart watches, cameras, locks
HS160930H	Thin square	16 × 9 × 3.0	8 Ω	0.8 / 1.0	93 dB @ 0.8 W / 3 cc	1000 Hz	3 cc sealed	Phones, small radios, watches
HS151125H	Thin square	15 × 11 × 2.5	8 Ω	1.0 / 1.2	95 dB @ 0.8 W / 1 cc	900 Hz	1 cc sealed	Phones, tablets
HS151130H	Thin square	15 × 11 × 3.0	8 Ω	1.0 / 1.2	95 dB @ 1.0 W / 1 cc	900 Hz	1 cc sealed	Phones, tablets
HS121722H	Thin square	12 × 17 × 2.2	7 Ω	1.0 / 1.2	95 dB @ 1.0 W / 1 cc	850 Hz	1 cc sealed	Phones, tablets
HS201623H	Thin square	20 × 16 × 2.3	7 Ω	1.0 / 1.2	96 dB @ 1.0 W / 1 cc	800 Hz	1 cc sealed	Phones, tablets
HS250926H	Thin square	25 × 9 × 2.6	4 Ω	2.0 / 2.5	93 dB @ 2.0 W / 2 cc	700 Hz	2 cc sealed	Phones, tablets, laptops
HS251233H	Thin square	25 × 12 × 3.3	4 Ω	2.0 / 2.5	97 dB @ 2.0 W / 1 cc	750 Hz	1 cc sealed	Phones, tablets, laptops
HS341135H	Thin square	34 × 11 × 3.5	4 Ω	2.0 / 2.5	98 dB @ 2.0 W / 3 cc	650 Hz	3 cc (F0 at 2 cc)	Rugged phones, industrial tablets
HS361331H	Thin square	36 × 13 × 3.1	4 Ω	2.0 / 2.5	99 dB @ 2.0 W / 4 cc	600 Hz	4 cc (F0 at 2 cc)	Industrial tablets, laptops, desk lamps
HS001342H42	Small round	φ13 × 4.2	8 Ω	0.5 / 0.8	88 dB @ 0.5 W	1000 Hz	Small sealed cavity	Cameras, toothbrushes
HS001534H39	Small round	φ15 × 3.4	8 Ω	0.8 / 1.0	90 dB @ 0.8 W	800 Hz	Small sealed cavity	Door locks, voice remotes
HS241540H42	Round	24 × 15 × 4.0	8 Ω	0.8 / 1.0	93 dB @ 0.8 W	800 Hz	Moderate cavity	Locks, tablets, alarms
HS241534H34	Round, dual magnet	24 × 15 × 3.4	8 Ω	1.0 / 1.2	95 dB @ 1.0 W	800 Hz	Moderate cavity	Same as 2415 when output is short
HS002038H	Round	φ20 × 3.8	8 Ω	0.8 / 1.0	93 dB @ 1.0 W	800 Hz	Moderate cavity	POS, consoles, alarms
HS002045H	Round	φ20 × 4.5	8 Ω	0.8 / 1.0	94 dB @ 1.0 W	600 Hz	Moderate cavity	POS, consoles, smart home
HS204130H30	Track	20 × 14 × 3.0	8 Ω	1.0 / 1.2	90 dB @ 1.0 W	900 Hz	Structural cavity	POS, toothbrushes, doorbells
HS280935H35	Track	28 × 9 × 3.5	8 Ω	1.0 / 1.2	91 dB @ 1.0 W	900 Hz	Structural cavity	Smart home, radios
HS-BX-1217-T26	1217 box, front fire	19 × 14 × 5.0	8 Ω	1.0 / 1.2	95 dB @ 1.0 W	1350 Hz	None — integral	Eye massagers, prompt only
HS-BX-1915	1217 box, side fire	19 × 15 × 5.5	8 Ω	1.0 / 1.2	95 dB @ 1.0 W	1150 Hz	None — integral	Voice products
HS-BX-2030	1217 box, side fire	20 × 30 × 4.0	8 Ω	1.0 / 1.2	95 dB @ 1.0 W	950 Hz	None — integral	Voice products
HS-BX-1217-3813X	1217 box, side fire	38 × 18 × 3.5	8 Ω	1.0 / 1.2	95 dB @ 1.0 W	850 Hz	None — integral	Tablets, photo frames, portable monitors

Reading the enclosure-needed column is the fastest way to shortlist. If the industrial design has already frozen the housing and no sealed cavity can be guaranteed, only the box-module rows are viable. If a cavity exists but its volume is under 1.5 cc, the 3 cc and 4 cc bare drivers will not deliver their published figures and should be dropped regardless of how good those figures look.

  5. Leading Compact Speaker Manufacturers Compared

Table 4: Compact loudspeaker suppliers — capability comparison (no pricing or capacity data)

Supplier type	Typical strength	What to verify	Best fit
Specialist micro-loudspeaker maker with in-house box modules (Recommended)	Bare drivers and integrated box modules from one source; cavity tuning done on the same bench that measures the driver	Ask for the F0 figure in *your* box volume, not the datasheet box	Programs where the cavity is still moving or under 2 cc
High-volume consumer cell maker	Tight unit-to-unit spread, mature automated lines, strong handset references	Minimum order quantities and change-control discipline	Very high volume, stable designs
Broad electroacoustic catalogue house	Wide footprint coverage, fast sample turnaround	Whether the published SPL states a test box	Early exploration, many size options
Regional assembly / trading partner	Local stock, low MOQ, short logistics	Traceability of the cell and who owns the acoustic design	Pilot builds and aftermarket replacement

  6. Objective Supplier Data Comparison

Table 5: What to request in an RFQ, and why each item matters for compact builds

Requested item	Why it matters in a compact device	Acceptable answer	Red flag
SPL and F0 measured in a stated box volume	The published figure is meaningless without the box	Both numbers with the volume named on the same sheet	"95 dB" with no box, no distance, no power
Unit-to-unit spread across a sample of 20+	Small cavities amplify part-to-part variation into audible inconsistency	A stated ±dB window at 2 kHz and at F0	Only a typical value, no spread
Sealed-box modules offered alongside bare drivers	Removes cavity design and leak risk from the host device	A module family sharing the same driver	Bare drivers only, cavity left to the customer
Operating temperature range and thermal test method	Compact sealed enclosures trap heat; power compression follows	A stated range plus the soak method	"Standard" with no numbers
Compliance documentation (RoHS / REACH)	Required for most consumer markets	Current declarations with dates	Declarations older than two years

Project Case Study — Buying Back 3.4 cc in a Handheld Terminal

A handheld terminal program had selected on sensitivity alone: a 36 × 13 × 3.1 mm bare driver publishing 99 dB at 2 kHz / 10 cm / 2.0 W into a 4 cc box — the loudest figure available in the compact class. On paper the design reserved 5.4 cc of total volume. In production it did not work out that way. The battery, the PCB standoffs and a cable route intruded into the reserved cavity, and the effective sealed volume measured closer to 2.4 cc. The result was a part that never delivered its datasheet number: SPL at 0.5 m sat at 79.2 dB(A), and because the effective cavity varied with assembly torque, unit-to-unit spread reached ±3.2 dB — wide enough that two terminals on the same bench sounded like different products. Working with Shenzhen Hongsheng Electronic Industry Co. LTD, the program moved the acoustic boundary into the part instead of the housing: an HS-BX-1511-F20T dual-box module, publishing 98 dB at 1.0 W with F0 at 880 Hz, carries its own enclosure, so the host reserves only the module outline. Reserved volume fell from 5.4 cc to roughly 2.0 cc. Measured SPL at 0.5 m landed at 78.4 dB(A) — 0.8 dB down, below the threshold anyone noticed in use — while unit-to-unit spread tightened to ±0.9 dB because there was no longer a gasket-dependent cavity to assemble inconsistently. The 3.4 cc released went into a battery roughly 12 percent larger.

The 99 dB part was never a 99 dB solution. It was a 2.4 cc solution wearing a 4 cc label.

  7. Selection Pitfalls to Avoid

4. Filtering the catalog by driver thickness or outline volume. The smallest part frequently carries the largest back-volume requirement. Rank by total displaced volume instead.

5. Comparing a 2.83 Vrms figure against a 1 W figure. On a 4 Ω part the voltage basis is about 2 W and roughly 3 dB hotter — normalise before ranking.

6. Reading a coupler-class figure as a loudspeaker SPL. HS080923H's 123 dB is a receiver measurement into a coupler and belongs to a different scale entirely.

7. Assuming the reserved cavity survives to production. Battery, standoffs and cabling routinely consume a third of it. Measure the effective volume in the first article-inspection build, not in CAD.

8. Ignoring the F0 box when there are two. HS341135H and HS361331H quote SPL at one volume and F0 at a smaller one; both are real, and conflating them produces a wrong cavity target.

9. Specifying a bare driver when the housing cannot be sealed. A gasket-dependent cavity in a snap-fit consumer housing is a leak waiting to happen; an integrated box module removes the variable.

10. Sizing the grille last. A compact outlet under about 2 percent open area can cost 4–6 dB in the 2–4 kHz speech band, which is precisely where the compact parts can least afford to lose output.

  8. Applicable Standards and Certifications

Table 6: Product and system-level standards relevant to compact consumer devices

Standard	Title	Relevance to this product
IEC 62368-1 / UL 62368-1	Audio/video, information and communication technology equipment — safety	Baseline safety for the host consumer device; applies to the finished product, not the driver alone
IEC 61000-6-1 / -6-3	Electromagnetic compatibility — generic immunity and emission standards	EMC behaviour of the finished device including its audio stage
CISPR 32 / EN 55032	EMC of multimedia equipment — emission requirements	Conducted and radiated emissions for consumer electronics placed on the EU market
IEC 60529	Degrees of protection provided by enclosures (IP code)	Relevant when the compact device claims any ingress rating; the grille is part of the boundary
RoHS Directive 2011/65/EU + 2015/863	Restriction of hazardous substances	Material compliance for the driver and its terminations

Table 7: Transducer-level test methods for the loudspeaker itself

Standard	Title	Test focus
IEC 60268-5	Sound system equipment — Part 5: Loudspeakers	Rated impedance, sensitivity, frequency response, distortion — the reference for how the SPL figure should be stated
IEC 60068-2 series	Environmental testing	Dry heat, damp heat, cold, thermal shock and vibration — the series behind most reliability claims
IEC 60068-2-2	Environmental testing — Test B: Dry heat	High-temperature soak used to check F0 and SPL drift in a sealed compact enclosure
IEC 60068-2-6	Environmental testing — Test Fc: Vibration (sinusoidal)	Mechanical robustness of the diaphragm, surround and termination
UL 94 / IEC 60695-11-10	Flammability of plastic materials	Flammability class of the diaphragm, frame and box-module housing
REACH (EC) 1907/2006	Registration, Evaluation, Authorisation and Restriction of Chemicals	Substance declarations for materials and adhesives

Standard numbers are cited as the currently published designations and are subject to the latest published version and to the product datasheet. Confirm the applicable edition with your certification body before locking a test plan.

How much back volume does a compact internal speaker actually need?

Read it off the datasheet rather than guessing: the sensitivity line states the box the figure was measured in. In this class the declared values cluster at 1 cc, 2 cc, 3 cc and 4 cc. If the housing cannot deliver the stated volume, the published SPL and F0 will not be met — output below resonance drops at roughly 12 dB per octave once the cavity is starved.

Why does the same driver sound quieter in my device than on the bench?

Nine times out of ten the effective cavity is smaller than the datasheet box, or it leaks. Both raise the air-spring stiffness: resonance climbs and the low end falls away. Measure the assembled device at 0.5 m and compare the curve against the datasheet curve — a lifted F0 and a steeper slope below it is the signature of a cavity that is too small or not sealed.

Is a box module always the more compact choice?

Not always, but usually when the housing cannot be sealed. HS-BX-1217-T26 displaces 1.33 cc total against HS151125H's 1.41 cc, so the totals are nearly identical — the difference is that the module needs no host cavity, no gasket and no leak check. Against a driver needing 3–4 cc, the module wins outright.

What does "dual magnet" or "dual box" buy me?

Output in the same footprint. HS241534H34 is 2 dB louder than the standard HS241540H42 while being 0.6 mm thinner, and in the box families HS-BX-1511-F20T reaches 98 dB against HS-BX-1511-HLX01's 95 dB with a lower F0 as well (880 Hz against 950 Hz). It is the move to try when the housing is frozen.

Should I pick 4 Ω or 8 Ω?

At a fixed supply voltage, 4 Ω draws roughly twice the current and about twice the power — around 3 dB more output. The cost is amplifier current capability and battery runtime. Mains-powered devices with a reasonable Class-D stage usually take 4 Ω; battery wearables and sensors usually stay at 8 Ω.

How much grille open area is enough?

Treat 5 percent open area over the speech band as a working floor. Below about 2 percent, losses of 4–6 dB in the 2–4 kHz region are common, and that is the band compact drivers can least afford to lose. Increase hole count before increasing hole size, so the visual appearance stays under the industrial designer's control.

More in This Series — Compact Internal Speakers for Consumer Electronics

This article is Part 1 of a three-part technical series on compact internal speakers for consumer electronics. The other two parts cover the enclosure-volume engineering in detail, and the symptom-level troubleshooting of compact builds.

· Part 2 — Technical Requirements: How Much Enclosure Volume You Actually Need →https://www.hsdz-spk.com/news/524.html

· Part 3 — FAQ: Thin Sound, Rattle and Lost Output →https://www.hsdz-spk.com/news/525.html

  9. Summary — Choosing the Right Compact Internal Speaker

Start by converting every candidate into total displaced volume: driver outline plus the back volume its own datasheet declares. That single step reorders the shortlist, because the loudest parts in this class are consistently the most expensive in cubic centimetres. Decide next whether the housing can genuinely deliver a sealed cavity of the required size through production — if it cannot, an integrated box module removes the variable and often costs less volume overall. Where the cavity exists but is tight, stay in the 1–2 cc band, where the compact range returns its best output per cubic centimetre. Choose suppliers who publish both figures on one sheet, state a unit-to-unit spread rather than a typical value, and can measure the part in your box rather than only in theirs.