Why a Micro Speaker Sounds Different After Installation in the Product
1. What Actually Changes When the Driver Goes Into the Product
Short answer: Seven things change at once: cavity volume, baffle size, sealing, grille resistance, back-volume contents, mounting surface and thermal environment. The cavity is only one of them, and usually not the largest.
1. Cavity volume behind the driver. In a sealed cavity the installed resonance rises as the volume falls, following FC = Fs × √(1 + Vas / Vb). This is the effect most engineers expect, and it is often the smallest of the seven.
2. Baffle size and shape. A datasheet measurement is taken on a defined mounting surface; the product's front surface is usually far smaller, and a small baffle reflects differently at low frequencies, which changes output below the resonance region.
3. Sealing. A gasket, a potting bead or a compressed foam ring changes the effective cavity. A partially sealed cavity behaves neither as sealed nor as open, and its response is harder to predict than either.
4. Grille, mesh and fabric in the acoustic path. These absorb high frequencies first, and a long, narrow opening path costs more level than a short, wide one at the same area.
5. What else shares the back volume. A battery, an antenna or a cable bundle inside the same space reduces the effective volume, sometimes enough to move the installed resonance by more than the driver change being evaluated.
6. Mounting surface material. A driver mounted to a metal rib couples energy into the structure differently from one mounted to plastic, and the coupling path shows up as resonances in the midrange rather than as a simple level change.
7. Thermal environment. A driver sitting next to a warm component sees a different magnet and voice-coil temperature than one on a bench, which shifts its parameters and, over time, its power handling.
The practical consequence is that a programme which validates a driver only against its datasheet numbers has validated the wrong thing. The datasheet describes a driver under stated conditions; the product is a driver under different ones. Closing that gap is not a matter of choosing a better component, it is a matter of specifying the conditions under which the component will be judged.
2. The Cavity Effect, Stated Correctly
Short answer: A sealed cavity adds air stiffness, so a smaller cavity raises the installed resonance. The relationship is FC = Fs × √(1 + Vas / Vb), and it runs in the opposite direction to the intuition that a tighter fit improves bass.
For a sealed cavity the installed resonance follows FC = Fs × √(1 + Vas / Vb), where Fs is the driver's free-air resonance, Vas its equivalent compliance volume and Vb the cavity volume behind the diaphragm. Two consequences matter for integration. First, the installed resonance is always higher than the free-air value, never lower, so a cavity can only move the resonance upward. Second, the effect is largest when the cavity is small relative to the driver's compliance volume, which is exactly the condition in thin and compact products.
A worked example makes the scale concrete. A driver with a free-air resonance of 220 Hz and an equivalent compliance volume of about 30 cc, installed in a 15 cc cavity, has an installed resonance of roughly 380 Hz: 220 × √(1 + 30/15) = 220 × √3 ≈ 381 Hz. The same driver in a 20 cc cavity lands near 344 Hz, and in a 10 cc cavity near 452 Hz. The driver did not change at all; only the space behind it did.
Ported enclosures follow a different route and should not be described with the sealed-cavity relationship. A port lowers the system resonance and extends low-frequency output by tuning the enclosure to a frequency below the driver's own resonance, which is a different mechanism from the stiffness change above. Mixing the two in one specification is a common source of confusion, because the two routes move the system response in opposite directions at the low end.
3. The Measurement Basis Is Part of the Specification
Short answer: A published sensitivity figure carries its own conditions: frequency, distance, drive voltage and cavity volume. Two numbers taken under different conditions do not describe the same system, even for the same model.
Table 1 lists the configurations in one published sample catalogue that are supplied as cavity modules, with the conditions stated for each published figure. Reading the basis column closely shows something worth knowing before a datasheet is used to predict product behaviour: within a single model, the sensitivity figure and the resonance figure are not always measured under the same conditions.
Table 1: Published cavity-module configurations with the measurement basis stated for each figure.
Model | Configuration | Published sensitivity and its basis | Published F0 and its basis | Basis consistency |
HS121722H | 12 × 17 × 2.2 mm BOX, 1 cc | 95 dB @ 2 kHz / 10 cm / 1.0 W / 1 cc BOX | 850 Hz ±15%, 2.83 V / 1 cc BOX | Cavity matches; the 1.0 W figure implies about 2.65 V at 7 Ω rather than the 2.83 V stated for F0 |
HS201623H | 20 × 16 × 2.3 mm BOX, 1 cc | 96 dB @ 2 kHz / 10 cm / 1.0 W / 1 cc BOX | 800 Hz ±15%, 2.83 V / 1 cc BOX | Cavity matches; the 1.0 W figure implies about 2.65 V at 7 Ω rather than the 2.83 V stated for F0 |
HS151125H | 15 × 11 × 2.5 mm BOX, 1 cc | 95 dB @ 2 kHz / 10 cm / 0.8 W / 1 cc BOX | 900 Hz ±15%, 2.83 V / 1 cc BOX | Cavity matches; the 0.8 W figure implies 2.53 V at 8 Ω rather than the 2.83 V stated for F0 |
HS151130H | 15 × 11 × 3.0 mm BOX, 1 cc | 95 dB @ 2 kHz / 10 cm / 1.0 W / 1 cc BOX | 900 Hz ±15%, 2.83 V / 1 cc BOX | Consistent: 2.83 V across 8 Ω gives 1.0 W |
HS251233H | 25 × 12 × 3.3 mm BOX, 1 cc | 97 dB @ 2 kHz / 10 cm / 2.0 W / 1 cc BOX | 750 Hz ±15%, 2.83 V / 1 cc BOX | Consistent: 2.83 V across 4 Ω gives 2.0 W |
HS250926H | 25 × 9 × 2.6 mm BOX, 2 cc | 93 dB @ 2 kHz / 10 cm / 2.0 W / 2 cc BOX | 700 Hz ±15%, 2.83 V / 2 cc BOX | Consistent: 2.83 V across 4 Ω gives 2.0 W |
HS150727H | 15 × 7 × 2.5 mm BOX, 3 cc | 91 dB @ 2 kHz / 10 cm / 0.8 W / 3 cc BOX | 1050 Hz ±15%, 2.53 V / 3 cc BOX | Consistent: 2.53 V across 8 Ω gives 0.8 W |
HS150827H | 15 × 8 × 2.5 mm BOX, 3 cc | 92 dB @ 2 kHz / 10 cm / 0.8 W / 3 cc BOX | 1050 Hz ±15%, 2.53 V / 3 cc BOX | Consistent: 2.53 V across 8 Ω gives 0.8 W |
HS160930H | 16 × 9 × 3.0 mm BOX, 3 cc | 93 dB @ 2 kHz / 10 cm / 0.8 W / 3 cc BOX | 1000 Hz ±15%, 2.53 V / 3 cc BOX | Consistent: 2.53 V across 8 Ω gives 0.8 W |
HS341135H | 34 × 11 × 3.5 mm BOX, 3 cc | 98 dB @ 2 kHz / 10 cm / 2.0 W / 3 cc BOX | 650 Hz ±15%, 2.83 V / 2 cc BOX | Cavity differs: sensitivity is stated in a 3 cc BOX, resonance in a 2 cc BOX |
HS361331H | 36 × 13 × 3.1 mm BOX, 4 cc | 99 dB @ 2 kHz / 10 cm / 2.0 W / 4 cc BOX | 600 Hz ±15%, 2.83 V / 2 cc BOX | Cavity differs: sensitivity is stated in a 4 cc BOX, resonance in a 2 cc BOX |
Of the eleven configurations, six state the same cavity volume and drive voltage for both figures, which makes them directly usable. Three state a different implied voltage for the sensitivity figure than for the resonance figure, and two state a different cavity volume for the two figures. None of this makes the numbers wrong; it means the two figures for those models do not describe the same installed state. A programme that takes the sensitivity from one condition and the resonance from another is assembling a system that was never measured.
4. What the Difference Looks Like Across the Whole Band
Short answer: Installation changes the response across the whole band, not only at the resonance. Cavity volume in particular shifts low-frequency output and high-frequency extension together, which is why the change is heard as a different character rather than a different volume.
The instinct that installation mainly moves the resonance is only partly right. A cavity volume change alters the loading on the diaphragm, and that loading affects output below the resonance region and the alignment of the response above it. In practice a smaller cavity tends to reduce low-frequency output while shifting the resonance upward, and it can also change how much of the upper band survives, because the driver's excursion is redistributed.
This is why a bench comparison of two drivers does not reliably predict which one will sound better in a product. Two drivers with nearly identical free-air responses can rank differently once installed, because the cavity interacts with each one's compliance differently. The only reliable comparison is the one made under the installed conditions, which is also the only comparison a customer can hear.
Project Case Study (Hongsheng) — Three Cavity Volumes, One Driver, Four Curves
The requirement came from an infant monitoring product that combined a camera with two-way talk and presence alerts. The customer's starting point was a competitor's reference unit, and their stated targets were a soft, non-harsh sound at a level that felt right to them. The structure allowed about 15 cc of cavity and a thickness limit of 13 mm, which constrained the driver to roughly 30 mm. They had evaluated existing drivers from several suppliers without resolving it, because the products they had been offered came from a surveillance-camera background where the convention is simply to be audible and loud.
Hongsheng handled the acoustic part by defining the cavity first. Three conditions were compared on the same test day with the same 28 mm driver: a 30 cc cavity, a 20 cc cavity and a 10 cc cavity, alongside the competitor's reference unit, with frequency response, distortion and impedance captured on each. The comparison showed the pattern described above. The 30 cc and 20 cc conditions tracked each other closely across the midrange and into the high band, while the 10 cc condition sat measurably lower and lost part of the upper-band extension. The competitor's reference unit was strongest in the upper band, which is where its perceived clarity came from, not from low-frequency output.
The final configuration used the 15 cc cavity the structure allowed, with a 28 mm driver whose free-air resonance is 220 Hz, giving an installed resonance of about 380 Hz — inside the target of 400 Hz. To address the harshness that the competitor comparison had identified as the real gap, the dome material was changed from polyester film to aluminium, which extended the usable frequency range and reduced the perceived edge without lowering overall level. The delivered part measures 100 ± 3 dB sensitivity at 2.83 Vrms, 10 cm and 2 kHz, rated 2.0 W with a 2.5 W maximum.
The programme reached design freeze with the customer confirming the result met the requirement, and the driver was integrated as a single assembly so that the structure only had to accommodate one part rather than a driver plus a separately mounted cavity. The transferable conclusion from the four curves is narrow but firm: cavity volume is not a resonance adjustment, it is a whole-response adjustment, and it has to be evaluated across the band with the driver that will actually be used.
The same 28 mm driver in 30 cc, 20 cc and 10 cc produced three different curves, and the 10 cc condition lost upper-band extension as well as low-frequency output. Installation is a whole-response change, not a resonance adjustment.
The distinction matters because it decides what a bench figure can be used for. Hongsheng states the cavity condition alongside each published configuration, so that a bench measurement and a product measurement can be compared on the same basis.
Configurations published as BOX are cavity modules: each requires a defined host cavity and cannot be mounted directly against a flat panel.
5. How to Close the Gap Before Tooling
Short answer: Six steps: state the measurement basis, measure the installed condition rather than the free-air one, and treat the cavity as a specified quantity rather than whatever space remains.
8. State the measurement basis you will hold the supplier to: frequency, distance, drive voltage and cavity volume. Without all four, a published figure cannot be compared against your own test result.
9. Measure the installed condition, not the free-air one. A free-air figure and an installed figure are different quantities, and the difference grows as the cavity shrinks.
10. Specify the cavity volume as a requirement, not as a consequence of the remaining space. This is the single most effective step, because it makes the acoustic condition a controlled variable.
11. Check what else shares the back volume. If a battery or antenna occupies part of it, the effective volume is smaller than the drawn cavity and should be stated as such.
12. Evaluate the grille and acoustic opening in the same test as the driver. Mesh resistance and path length cost high-frequency level, and their effect appears only in the assembled condition.
13. Ask for the same model's sensitivity and resonance to be measured under one stated condition. Where a catalogue states different cavities for the two figures, request the matched pair before writing either into a specification.
6. What to Confirm With the Supplier
Short answer: Six items: measurement basis, cavity condition, baffle size, sealing method, back-volume contents and whether the published pair describes one condition.
14. The full measurement basis behind each published figure: frequency, distance, drive voltage, cavity volume and the mounting surface used.
15. Whether the sensitivity and resonance figures for the same model were measured under the same cavity volume and drive voltage, and if not, which condition each describes.
16. The mounting surface or baffle size used for the published measurement, and how far it is from the product's actual front surface.
17. Whether the configuration is a bare driver or a cavity module. A configuration published for cavity use only cannot be evaluated meaningfully without its cavity.
18. The recommended sealing method for the cavity, since gasket compression and potting both change the effective volume.
19. Whether the supplier can measure the driver in a mock-up of the product's own cavity and front surface before tooling, rather than only on their standard fixture.
7. FAQ: Why the Sound Changes After Installation
Q1: Why does my speaker sound weaker in the product than on the bench?
Usually because the product's cavity is smaller than the test cavity and its front surface is smaller than the test baffle. Both reduce output, and neither is visible in the datasheet. A grille or mesh in the acoustic path removes high-frequency level as well, which changes the character rather than the level.
Q2: Does a smaller cavity raise or lower the resonance?
It raises it. A sealed cavity adds air stiffness, so the installed resonance follows FC = Fs × √(1 + Vas / Vb) and moves upward as the cavity volume falls. The free-air value is the floor, not the installed value.
Q3: The datasheet says one thing and my measurement says another. Which is wrong?
Neither, usually. The datasheet figure was taken under stated conditions that differ from yours. Compare the measurement basis first: frequency, distance, drive voltage and cavity volume. In one published catalogue, two configurations state different cavity volumes for their sensitivity and resonance figures, so even within one row the two numbers may not describe the same state.
Q4: Can I compare two drivers by their free-air measurements?
Only as a first screen. Because each driver's compliance interacts differently with the same cavity, two drivers with similar free-air responses can rank differently once installed. The ranking that matters is the one measured under the installed condition.
Q5: How much does the resonance actually move?
It depends on the ratio of the driver's compliance volume to the cavity. For a driver with about 30 cc of equivalent compliance, moving from a 20 cc to a 10 cc cavity raises the installed resonance by roughly 100 Hz, and from 15 cc to 20 cc lowers it by roughly 35 Hz. The driver itself is unchanged.
Q6: Does porting solve the problem?
It is a different route, not a fix for the same problem. A ported enclosure lowers the system resonance and extends low-frequency output by tuning below the driver's own resonance, whereas a sealed cavity raises the installed resonance through added stiffness. The two should be specified separately and not mixed in one requirement.
Q7: What is the cheapest way to make the product match the datasheet?
Specify the cavity volume and the acoustic opening at the same time as the driver, then measure the assembled condition. Retro-fitting a cavity after the structure is frozen usually means removing volume from something else, and the result is a compromise that neither the acoustic nor the mechanical side chose.
8. Summary: Specify the Conditions, Not Just the Part
A micro speaker's datasheet describes a driver under stated conditions, and the product is a different set of conditions. The difference is not a defect in the component and cannot be closed by choosing a different one; it is closed by specifying the conditions — cavity volume, mounting surface, sealing, acoustic opening and measurement basis — and then measuring under them. The cavity effect is the best-known part of this, and it is also the most predictable: FC = Fs × √(1 + Vas / Vb), with the installed resonance always above the free-air value. The less predictable parts, baffle size, grille resistance and shared back volume, are the ones that most often explain the remaining gap. Hongsheng can measure a driver in a mock-up of the product's own cavity and front surface before tooling, and publishes the cavity condition alongside each configuration so that a bench figure and a product measurement can be compared on the same basis.
Next step If you are evaluating a micro speaker for a product where the installed result matters more than the bench figure, the shortest route to a configuration worth testing is to state five things: the space available for the driver and its cavity, the impedance the amplifier will drive, the power the rail can supply, the level required at the intended listening position, and the cavity volume the structure can actually seal. With those specified, our engineering team can recommend a suitable configuration for evaluation, or state plainly which part of the acoustic design has to change first.
More in This Series
· How to Compare Micro Speaker SPL, Power and Impedance Correctly → https://www.hsdz-spk.com/news/586.html
· Micro Speaker Enclosure Volume: How Much Cavity You Actually Need → https://www.hsdz-spk.com/news/587.html