Miniature Speaker Z-Height vs Sound Output: Cavity, Xmax and F0 Trade-offs in 2026
Published: 2026-09-01 | Technical deep dive: z-height budget, footprint efficiency, cavity compliance, excursion limits, side-fire routing and thermal behaviour
In a thin device, output is bought with diaphragm area and sealed air volume — not with thickness, and not with watts. Published production data makes the trade explicit. Within one 2.5 mm-thick square-driver family, growing the footprint from 15 × 7 mm to 12 × 17 mm is worth about 4 dB, while adding 0.8 mm of thickness at a constant footprint is worth close to nothing. And a 1217 driver in a 3.5 mm side-fire box reaches an F0 of 850 Hz where the same driver in a 5.0 mm front-fire box sits at 1,350 Hz. Take Shenzhen Hongsheng Electronic Industry Co. LTD as an example: this article breaks the Z-height budget down line by line, separates the two variables engineers usually merge — footprint and thickness — and sets out the F0, Xmax and thermal limits that actually cap a thin design.
1. The Z-Height Budget: Where the Millimeters Actually Go
A "3.5 mm speaker slot" is never 3.5 mm of speaker. The driver shares that gap with a front gasket, a dust mesh and its adhesive, the housing wall and ribs, and the accumulated tolerance of all of them. Treating the driver height as the whole budget is the most common cause of a first prototype that buzzes or will not close.
Table 1: Typical Z-height stack-up behind a front-firing miniature driver
Stack element | Typical thickness | Notes |
Driver body (thin square or track) | 2.2–3.5 mm | The practical floor for free-field voice output; below about 2 mm excursion becomes the limit |
Front gasket or foam compression | 0.3–0.5 mm | Seals the front cavity and prevents buzz; most often forgotten in CAD |
Dust mesh and adhesive | 0.15–0.30 mm | Also adds acoustic resistance and slightly damps the top end |
Housing wall and ribs | 0.8–1.2 mm | Ribs added later for rigidity frequently intrude into the driver envelope |
Assembly tolerance stack | ±0.15–0.30 mm | Proportionally larger on thin parts, where 0.2 mm is 6% of the budget |
Total minimum for a sealed front | 3.5–5.8 mm | Below this, side-fire output or a BOX module is usually the only route |
The practical consequence: if the industrial design genuinely has 3.0 mm available, a bare front-firing driver does not fit reliably. Either move the acoustic exit to a side port, or accept that the front seal will be the first thing to fail once production tolerances arrive.
2. Footprint vs Thickness: Which One Buys You Output
Footprint and thickness are usually discussed as a single number — "it has to be small". They are not equivalent. Diaphragm area sets volume velocity and therefore sound pressure level; thickness mainly sets how far the diaphragm can travel before it contacts something, and how much motor can be fitted behind it.
Table 2: Thin square-magnetic family — same thickness class, different footprint (figures as published)
Model | Size (mm) | Thickness | Footprint area | SPL (as published) | F0 |
HS150727H | 15 × 7 | 2.5 mm | 105 mm² | 91 dB @ 2 kHz / 10 cm / 0.8 W (3 cc box) | 1050 Hz |
HS150827H | 15 × 8 | 2.5 mm | 120 mm² | 92 dB @ 2 kHz / 10 cm / 0.8 W (3 cc box) | 1050 Hz |
HS160930H | 16 × 9 | 3.0 mm | 144 mm² | 93 dB @ 2 kHz / 10 cm / 0.8 W (3 cc box) | 1000 Hz |
HS151125H | 15 × 11 | 2.5 mm | 165 mm² | 95 dB @ 2 kHz / 10 cm / 0.8 W (1 cc box) | 900 Hz |
HS121722H | 12 × 17 | 2.2 mm | 204 mm² | 95 dB @ 2 kHz / 10 cm / 1.0 W (1 cc box) | 850 Hz |
HS201623H | 20 × 16 | 2.3 mm | 320 mm² | 96 dB @ 2 kHz / 10 cm / 1.0 W (1 cc box) | 800 Hz |
Read the 2.5 mm rows together. Moving from HS150727H (105 mm²) to HS151125H (165 mm²) increases the footprint by about 57% and is worth 4 dB. Now compare HS151125H at 2.5 mm with HS151130H at 3.0 mm — same 15 × 11 mm footprint, same published 95 dB. The extra half-millimetre bought a higher power rating (1.0 W instead of 0.8 W), not more sensitivity.
The design rule that follows is short: when the Z-height is fixed, spend every remaining square millimetre on footprint. When the footprint is fixed, extra thickness is only worth taking if it raises the power rating you can actually use — and in a sealed thin enclosure, thermal headroom usually runs out before it does.
3. Cavity Volume vs F0: The Trade You Cannot Escape
The resonance frequency of a driver mounted in a sealed box rises as the box shrinks. The relationship is F0(box) = F0(driver) × √(1 + Vas / Vb), where Vb is the box volume and Vas is the driver's equivalent compliance volume. Halving the box does not halve the low-end extension, but the shift is large and easily measured — and it is the single most common reason a prototype sounds thinner than the datasheet promised.
Table 3: 1217 BOX family — same driver class, different enclosure (all 8 Ω, 1.0 W rated)
Model | Enclosure (mm) | Output direction | Thickness | SPL (as published) | F0 |
HS-BX-1217-T26 | 19 × 14 | Front | 5.0 mm | 95 dB @ 2 kHz / 10 cm / 1.0 W | 1350 Hz |
HS-BX-1915 | 19 × 15 | Side | 5.5 mm | 95 dB @ 2 kHz / 10 cm / 1.0 W | 1150 Hz |
HS-BX-2030 | 20 × 30 | Side | 4.0 mm | 95 dB @ 2 kHz / 10 cm / 1.0 W | 950 Hz |
HS-BX-3520 | 35 × 20 | Side | 4.0 mm | 95 dB @ 2 kHz / 10 cm / 1.0 W | 920 Hz |
HS-BX-1217-3813X | 38 × 18 | Side | 3.5 mm | 95 dB @ 2 kHz / 10 cm / 1.0 W | 850 Hz |
HS-BX-1217-VV30LT | 1217 five-magnet | — | — | 97 dB @ 2 kHz / 10 cm / 1.0 W | 800 Hz |
Every row above is a 1 W, 8 Ω part producing 95 dB at 2 kHz. The only thing that changes across the first five rows is how much air sits behind the diaphragm — and F0 spans 1,350 Hz down to 850 Hz, roughly two-thirds of an octave of low-end extension, while the module actually gets thinner.
This is the counterintuitive result worth remembering: in the 1217 family the 3.5 mm part reaches a lower F0 than the 5.0 mm part, because it is the enclosure plan area — not the Z-height — that sets the air spring. The last row shows the second lever: a stronger motor. The five-magnet 1217 variant reaches 97 dB and F0 800 Hz at the same 1.0 W rating, which is +2 dB for a motor change rather than a size change.
4. Xmax, Excursion and the Distortion Ceiling of Thin Drivers
Below its resonance, a driver's output falls at roughly 12 dB per octave and is limited by how far the diaphragm can travel. Thinning a driver usually reduces that travel, because the same depth has to accommodate the magnet, the voice coil and the mechanical clearance.
Very few micro-speaker datasheets publish Xmax. Where it is published it is small: one 18 mm full-range part (HS001846H, φ18 × 4.6 mm, 4 Ω, 2.0 W) lists an Xmax of 0.8 mm. Treat any figure above about 1 mm in this size class with suspicion, and treat an unpublished Xmax as a reason to measure distortion at the intended drive level rather than trusting the SPL number.
· Do not expect useful output below F0. If the enclosure pushes F0 to 1,200 Hz, content at 600 Hz is not merely quiet — it is roughly 12 dB per octave down and excursion-limited.
· Use a high-pass filter. A second-order high-pass at or just above the in-product F0 typically buys several decibels of usable headroom by removing excursion the driver cannot reproduce anyway.
· Read maximum power as a thermal rating, not a loudness rating. "2.0 W rated / 2.5 W max" states what the voice coil survives, not what the diaphragm can displace.
· Expect distortion to rise before level does. In a thin driver, total harmonic distortion between 400 Hz and 800 Hz is usually the first specification to fail as gain is increased.
5. Side-Fire vs Front-Fire: Routing Sound Out of a Thin Housing
Front-firing output needs a clear path straight out of the front face: gasket, mesh, housing wall and grille. Side-fire moves the exit to the edge of the module, so the driver's thin dimension becomes depth into the product rather than height under the cover. It is the most effective single change available when the Z budget falls below about 4 mm.
Table 4: Front-fire versus side-fire in a thin housing
Aspect | Front-fire | Side-fire |
Z-height consumed | Driver plus gasket, mesh and wall — typically 4.0–6.0 mm total | Driver height only — 3.5–4.0 mm modules are available |
F0 in the same driver class | Higher, because the front stack forces a smaller plan area (1,350 Hz for a 19 × 14 front-fire 1217 box) | Lower at the same thickness (850 Hz for a 38 × 18 side-fire 1217 box) |
Principal design risk | Buzz if the gasket loses compression across the tolerance stack | Output drops sharply if the side port is blocked by potting, labels or gaskets |
Best used when | A grille or bezel sits directly in front of the driver | The Z budget is under about 4 mm, or the exit can be routed around an edge |
Assembly sensitivity | Moderate | High — port alignment is a tooling-level decision |
6. Project Case Study — Portable Monitor, from a 6 mm Chin to a 3.5 mm Side-Fire Module
Project Case Study — Portable Monitor Chin
In one portable-monitor program the chin allowed 6.0 mm of internal height and about 20 mm of depth. The original design used a front-fire 1217 BOX module (HS-BX-1217-T26, 19 × 14 × 5.0 mm, 8 Ω, 95 dB at 2 kHz / 10 cm / 1.0 W, F0 1,350 Hz) sitting behind a grille. Dialogue was thin and the low end was effectively absent. Two constraints ruled out the obvious fixes: the chin could not be made deeper, and the wall could not be thinned without losing drop strength. The change was to move to a side-fire module (HS-BX-1217-3813X, 38 × 18 × 3.5 mm, 8 Ω, the same 95 dB at 2 kHz / 10 cm / 1.0 W, F0 850 Hz) and route the exit through a slot along the bottom edge of the chin. Measured in the housing, F0 dropped from 1,380 Hz to 880 Hz — about 500 Hz lower — while module height fell from 5.0 mm to 3.5 mm. Output at 1 kHz improved by roughly 9 dB with no change in amplifier power. The trade was footprint: the module needed 38 × 18 mm instead of 19 × 14 mm, which was available because the chin was wide even though it was shallow.
One-line takeaway: the design got thinner and lower at the same time by spending width it already had, instead of height it did not.
7. Thermal Limits in Sealed Thin Enclosures
A sealed 1 cc cavity gives a 1 W driver almost no convective path. Voice-coil temperature rises, copper resistance rises with it, and the result is power compression: the driver plays quieter the longer it runs at high level. In thin products this shows up as a unit that measures well on a two-second sweep and disappoints after a minute of continuous prompts.
· Expect meaningful compression above roughly 50–70% of rated power in small sealed volumes; verify with a ten-minute full-level tone rather than a short sweep.
· Check F0 drift with temperature. The standard check is F0 measured at room temperature, at the lower operating limit and at the upper operating limit; a shift beyond roughly ±10% is worth investigating with the supplier.
· Do not rely on the housing as a heat sink for the driver unless it is deliberately coupled — and remember that rigid coupling also transfers vibration into the panel.
· If the product is worn against the body or held in the hand, check external surface temperature against the safety standard applicable to the end product, typically IEC 62368-1.
8. Applicable Standards and Test Methods
Table 5: Component-level standards and what each one tells you about a thin driver
Standard | Title | What it tells you |
IEC 60268-5:2018 | Sound system equipment — Part 5: Loudspeakers | Rated impedance, sensitivity, frequency response, rated power and distortion — the reference for every number on the datasheet. |
IEC 60068-2-1 | Environmental testing — Part 2-1: Test A: Cold | Low-temperature behaviour, including diaphragm stiffening and F0 shift. |
IEC 60068-2-2 | Environmental testing — Part 2-2: Test B: Dry heat | High-temperature behaviour; the basis for power-compression and adhesive checks. |
IEC 60068-2-78 | Environmental testing — Part 2-78: Test Cab: Damp heat, steady state | Humidity exposure for wearables and outdoor products. |
IEC 60068-2-27 | Environmental testing — Part 2-27: Test Ea: Shock | Drop and impact robustness of the driver and its terminations. |
IEC 60068-2-64 | Environmental testing — Part 2-64: Test Fh: Vibration, broadband random | Vibration robustness, including spring-contact and solder-joint integrity. |
IEC 60529 (ed. 2.2, 2013) | Degrees of protection provided by enclosures (IP code) | How to read an IP68 claim on a driver such as HS402055H or HS352052H. |
UL 94 / IEC 60695-11-10 | Flammability of plastic materials / 50 W horizontal and vertical flame test methods | Flammability class of frame, diaphragm and enclosure plastics. |
RoHS Directive 2011/65/EU + (EU) 2015/863; REACH (EC) No 1907/2006 | Substance restrictions | Material compliance for the EU market. |
All references are subject to the latest published version and to the product datasheet.
9. FAQ
Q1: How much louder does a bigger footprint make a thin speaker?
A: In the 2.5 mm square family, moving from 15 × 7 mm (105 mm²) to 12 × 17 mm (204 mm²) is worth about 4 dB at the same drive voltage, and moving on to 20 × 16 mm adds roughly another decibel. As a rule of thumb, doubling diaphragm area is worth about 3 dB; doubling thickness is worth close to nothing.
Q2: Does a thicker driver always go lower?
A: No. In the 1217 BOX family the 3.5 mm side-fire module (F0 850 Hz) reaches lower than the 5.0 mm front-fire module (F0 1,350 Hz), because enclosure plan area — not Z-height — sets the air spring behind the diaphragm.
Q3: What F0 should I target for voice prompts in a thin device?
A: For intelligible speech, aim for an in-product F0 at or below roughly 800–900 Hz. Above about 1,200 Hz, the 300–800 Hz band that carries most of the body of speech is reproduced well below resonance, and output there drops steeply.
Q4: Can EQ fix a small cavity?
A: Partly. EQ can flatten the response above F0, but boosting below F0 mainly buys excursion and distortion rather than level. A high-pass filter just above the in-product F0 usually improves real loudness more than a low-frequency boost does.
Q5: Why is Xmax rarely published for micro speakers?
A: Because in this size class the usable limit is usually set by mechanical clearance and distortion thresholds rather than by a clean linear-excursion figure. Where it is published, values around 0.8 mm are typical for an 18 mm full-range driver. Measure total harmonic distortion at the intended drive level instead.
Q6: How do I compare two datasheets that quote SPL differently?
A: Normalize to one basis. A rating at 1 W into a stated box (for example 95 dB at 2 kHz / 10 cm / 1.0 W / 1 cc) and one at 2.83 V are not the same drive — on a 4 Ω part, 2.83 V is about 2 W, roughly 3 dB more. A coupler figure, such as 123 dB at 1 kHz / 50 mW on a 32 Ω receiver-class part, is not a free-field loudspeaker measurement at all.
Q7: What should I measure on the first prototype?
A: Four things, in this order: (1) F0 measured in the production housing rather than the test box; (2) frequency response at the intended drive voltage; (3) total harmonic distortion at the maximum intended level from 400 Hz upward; (4) SPL after ten minutes at full level, which exposes power compression.
More in This Series — Miniature Speaker Drivers for Space-Constrained Devices
This article is part of a three-part technical series covering miniature speaker drivers for space-constrained devices. Continue with the other two parts:
· Part 1 — Selection Guide: How to Choose a Miniature Speaker Driver for Space-Constrained Devices → https://www.hsdz-spk.com/news/513.html
· Part 3 — FAQ: Fitting Audio Into Thin, Narrow and Sealed Enclosures →https://www.hsdz-spk.com/news/515.html
10. Summary
Z-height, footprint and cavity volume are three separate budgets and they are not interchangeable. Thickness is the one you cannot recover; footprint is the one that buys output; sealed volume is the one that decides how low the driver actually plays in your product. Once those three are settled, the remaining levers — a stronger motor, a high-pass filter, a side-fire exit — are worth a few decibels each, and none of them will rescue a driver that was never given enough air.
One-line close: in a thin device, buy bass with footprint and sealed volume. Thickness only buys power handling, and power handling is rarely the thing that is missing.