Internal Speaker Mounting and Integration: Ports, Gaskets and Buzz-Free Assembly

Internal Speaker Mounting and Integration: Ports, Gaskets and Buzz-Free Assembly

Published: 2026-09-01  |  Mechanical and acoustic integration of an internal speaker — back volume, port and grille design, gasket sealing, buzz-free mounting, and how the housing shifts F0 and SPL.

Once an internal speaker is mounted inside a housing, the housing stops being a container and becomes part of the loudspeaker. The same driver can measure 3 dB apart in two enclosures that look identical, purely because of back-volume and gasket differences. Take **Shenzhen Hongsheng Electronic Industry Co. LTD** as an example: across the internal-speaker programs we support, the most expensive field failures are almost never a bad driver — they are a housing that was frozen before the acoustics were understood. This article covers the four decisions that decide whether the finished device sounds right: how much back volume to leave, how to size the port and grille, how to seal it, and how to mount it so it does not buzz.

  1. Why the Housing Is Half the Loudspeaker

A bare driver is an incomplete acoustic device. It has a diaphragm and a motor, but no defined air load behind it and no defined outlet in front of it. Both of those come from the product. The catalog F0 and SPL of a driver are measured in a specific test box; the moment the driver sits in your housing, those numbers move. Two consequences follow that surprise teams on their first internal-speaker build:

· The back volume sets F0. A larger sealed cavity behind the driver lowers the resonance; a smaller one raises it. This is the single biggest lever on low-end reach, and it costs nothing in parts.

· The port and grille set the top end. Below about 5% open area, a grille starts removing exactly the 2–4 kHz band where voice intelligibility lives. The driver did not change — the hole pattern did.

The practical rule: freeze the internal layout — battery, mainboard, back volume and port area — before selecting the driver, not after. A driver chosen against a provisional layout will be wrong the day the battery grows by 2 mm, and the wrongness shows up as a 3 dB hole in the response that no equalizer can fill.

  2. Back Volume: How Much Cavity You Actually Need

There is no single correct back volume; it depends on what the product must do. The ranges below are engineering-judgment starting points for a sealed cavity behind a small internal driver, and the final value should be confirmed against the product datasheet and a measurement of the actual enclosure.

Table 1: Typical sealed back-volume targets by product class

Product class	Typical sealed back volume	What it buys you	Caveat
Voice-prompt (locked enclosure)	0.5–2 cc behind the driver	Just enough to define F0; output is set by the driver	Leaks here raise F0 and drop output below it
Media tablet / portable monitor	2–6 cc	Some mid-bass reach; lower F0 than a voice part	Needs a low-F0 driver to be useful
Projector / all-in-one	6–20 cc, or ported	Real bass below ~300 Hz	A port adds a tuned resonance below F0
Music / conference unit	20 cc+ or a passive radiator	Output below 200 Hz	Usually a multimedia cavity module, not a bare driver

Note the asymmetry: you can always lower F0 by enlarging the cavity, but you cannot recover output the driver does not have. A voice-prompt part at 950 Hz F0 will never sound like a music part at 500 Hz F0, however large you make the box — the driver excursion is the floor.

  3. Port and Grille Design: Where Output Is Won or Lost

The sound outlet is an industrial-design decision, but it is also an acoustic one. The open area of the grille and the length of the port path decide how much of the driver's 2–4 kHz output actually reaches the listener.

Table 2: Grille open area vs. measured loss in the speech band

Grille open area	Typical loss at 2–4 kHz	Acceptable for	Fix if you are here
< 2%	4–6 dB	Voice prompts only, quiet rooms	Re-cut the hole pattern or thin the mesh
2–5%	2–4 dB	Voice prompts, moderate noise	Acceptable; watch the mesh behind it
5–10%	< 2 dB	Media products	No action needed
> 10%	< 1 dB	Music / high-SPL products	No action needed

The usual correction is not a louder driver — it is a re-designed hole pattern. Moving open area from roughly 1.5% to 5% typically recovers 2–3 dB in the 2–4 kHz band, which is precisely where prompt intelligibility sits. A fine mesh laminated behind a low-open-area grille compounds the loss, so the mesh and the hole pattern must be designed together.

  4. Gaskets and Seals: The Difference Between a Box and a Leak

A sealed back volume only stays sealed if the gasket actually seals. An unsealed back volume leaks, and a leaky back volume raises F0 and drops output below it — the same failure mode as too small a cavity, but inconsistent unit to unit.

Table 3: Seal approaches ranked by production consistency

Seal approach	F0 spread across a lot	Effort	Best for
No seal (relies on housing fit)	Unpredictable	None	Never for SPL-critical products
Hand-placed adhesive bead	±2 to ±3 dB	Low	Prototypes and one-off builds
Die-cut self-adhesive gasket, fixed thickness	±0.5 dB	Medium	Production voice products
Molded-in rib plus driver locating lip	±0.3 dB	High (tooling)	High-volume media products

The cheapest repeatable result is a die-cut self-adhesive gasket of a fixed 0.2–0.4 mm thickness, located by the driver's own flange. It removes the operator variable that makes hand-beading spread F0 by several dB across a lot.

  5. Achieving Buzz-Free Mounting

Buzz and rattle are mechanical, not electrical, and they appear at high SPL or after shock — exactly when the product is being demonstrated. Four mechanisms account for nearly all of them.

Table 4: Buzz and rattle causes, symptoms and fixes

Cause	Symptom	Fix
Housing wall flexes against the driver frame	Buzz at maximum prompt volume	Stiffen the wall, add a standoff, or use a box module that carries its own enclosure
Loose or uneven screw torque	Rattle on bass content	Specify torque, add thread-lock, use a 4-screw box with defined clamping
Diaphragm contacts the grille	Scrape or buzz at high SPL	Increase the standoff 0.5–1.0 mm; check grille flatness
Voice-coil rub after a drop	Distortion after shock	Add suspension travel, consider potting or a tracked-magnet construction

The structural fix is better than the band-aid. A driver bonded straight to a thin housing wall loads that wall with every excursion; the wall becomes part of the moving system and buzzes. A box module removes the wall from the acoustic load entirely, which is why the same nominal driver stops buzzing the moment it is moved into a sealed module.

  6. How the Housing Shifts F0 and SPL — Measurement Reality

The datasheet F0 is measured in the vendor's test box. In your housing it will differ, and the direction is predictable.

Table 5: Common housing changes and their acoustic effect

Change from the test condition	Effect on F0	Effect on SPL	Note
Larger sealed back volume	Lowers F0	Small change	Roughly -15 to -20% per doubling of volume (engineering judgment)
Smaller back volume	Raises F0	Small change	+15 to +30% is common
Air leak in the back volume	Raises F0	Drops below F0	+5 to +10% F0 and about -2 dB output
Port added and tuned	Adds a resonance below F0	Extends low end	Only helps if tuned to the target band
Grille open area reduced	Little direct effect	Drops in 2–4 kHz	The speech band, not the low end

This is why a bare-driver project needs an acoustic measurement of the actual enclosure before sign-off. Comparing your measurement to the catalog F0 without accounting for back volume is the most common source of a 'the driver does not match the datasheet' complaint that is actually a housing problem.

  7. Material and Process Choices: Adhesive, Magnet, Surround

Three component choices decide whether the acoustic result holds up over temperature and time. They are commonly specified at the quotation stage, not discovered in the field.

Table 6: Component choices and what they protect

Choice	What it protects	Trade-off	When it matters
Center adhesive grade (e.g. 4505 / 868K / AB)	F0 stability across temperature	High-temp grades cost more	Outdoor or in-vehicle products from -10 to 70 °C
Magnet grade (standard vs H / M)	SPL at high temperature	H/M magnets cost more	Products that sit in direct sun or near heat sources
Surround (foam vs cloth vs rubber)	Damping and service life	Foam cheapest, rubber longest life	Foam for cost; rubber for long duty cycles
Diaphragm (paper vs aluminium dome)	Stiffness and high-frequency output	Aluminium dome lifts HF but costs more	Media products needing 4 kHz+ output

On the temperature question, the goal most programs quote is holding F0 within about ±10 Hz from -10 °C to 70 °C. That is achievable with a high-temperature center adhesive and an H/M-grade magnet, but it is a specification to confirm per part number, not a catalog default. Operating temperature range, THD and IP rating are not published in the standard catalog and should be treated as project-specific.

  8. Representative Internal Speaker Models by Mounting Type

The table collects production models grouped by how they are mounted and terminated. SPL is quoted as published; the measurement basis is noted where the catalog states it. Several thin square parts are marked 'for BOX use only' — they have no front cover and must be installed in a box you supply.

Table 7: Internal speaker models by mounting and termination

Model	Mounting / termination	Size (mm)	Imp.	Power	SPL	F0	Note
HS-BX-1511-HLX01	Box module, self-contained	1511 BOX	8 Ω	1.0 / 1.2 W	95 dB	950 Hz	AI robots, story machines
HS-BX-1217-3813X	Box module, side fire	38 × 18 × 3.5	8 Ω	1.0 / 1.2 W	95 dB	850 Hz	Tablets, monitors, photo frames
HS-BX-2512-QX01	Box module, self-contained	2512 BOX	4 Ω	2.0 / 2.5 W	97 dB	800 Hz	Pet feeders, tablets, laptops
HS-BX-3613-UDP01	Box module, self-contained	3613 BOX	4 Ω	2.0 / 2.5 W	98 dB	600 Hz	Smart desk lamps
HS-BX-282813H	Box module, φ15.5 motor	28 × 28 × 13	4 Ω	2.0 / 2.5 W	97 dB	880 Hz	AI voice products
HS-BX-203008H	Box module, ported, φ12.5	2030 BOX	4 Ω	3.0 / 4.0 W	96 dB	1000 Hz	All-in-one, industrial control
HS-BX-703017H	Box module, 4 screw holes	70 × 30 × 17	4 Ω	2.0 / 2.5 W	98 dB	850 Hz	Higher sound-quality voice
HS-BX-703314H	Box module, 2 screw holes	70 × 33 × 14.7	4 Ω	2.0 / 2.5 W	98 dB	700 Hz	Higher sound-quality voice
HS003050H	Bare driver, with cover, solder	φ30 × 5.0	8 Ω	2.0 / 2.5 W	97 dB	550 Hz	Robots, POS, IoT voice
HS003650H	Bare driver, with cover, solder	φ36 × 5.0	8 Ω	2.0 / 2.5 W	97 dB	500 Hz	Robots, surveillance, IoT
HS241534H34	Bare driver, dual magnet, solder	24 × 15 × 3.4	8 Ω	1.0 / 1.2 W	95 dB	800 Hz	Door locks, tablets, alarms
HS284011H	Bare driver, spider, solder	28 × 40 × 11	4 Ω	3.0 / 4.0 W	95 dB	500 Hz	Smart home, projectors
HS402055H	Bare driver, IP68, track, wire	40 × 20 × 5.5	8 Ω	2.0 / 2.5 W	97 dB	570 Hz	Walkie-talkies, rugged phones
HS352052H	Bare driver, IP68, dual magnet	35 × 20 × 5.2	8 Ω	2.0 / 2.5 W	97 dB	650 Hz	Walkie-talkies, rugged phones
HS151125H	Bare driver, with cover, leaf spring	15 × 11 × 2.5	8 Ω	1.0 / 1.2 W	95 dB	900 Hz	Smartphones, tablets
HS121722H	Bare, no front cover, BOX use	12 × 17 × 2.2	7 Ω	1.0 / 1.2 W	95 dB	850 Hz	Smartphones, tablets (in a box)
HS201623H	Bare, no front cover, BOX use	20 × 16 × 2.3	7 Ω	1.0 / 1.2 W	96 dB	800 Hz	Smartphones, tablets (in a box)
HS080923H	Receiver, leaf spring, 32 Ω	8 × 9 × 2.3	32 Ω	50 / 80 mW	123 dB*	600 Hz	Coupler-class; *measured into a 2 cc coupler

The mounting column is the real differentiator here. A part such as HS-BX-703017H with four screw holes is built to be clamped to a panel and stay put; a part such as HS121722H with no front cover is built to live inside a box you supply. Specifying the second one as a panel-mounted part is the fastest way to a failed build.

Project Case Study — Smart-Home Panel: Grille and Gasket, Not the Driver

A smart-home control panel used a φ36 × 5.0 mm bare driver (HS003650H, 8 Ω, 2.0 W rated, 97 dB, F0 500 Hz) mounted behind a punched-metal grille, with a hand-dispensed silicone bead as the back-volume seal. Bench units sounded acceptable, but the production line told a different story: measured intelligibility at 1 m averaged only 68 dB(A), and the 3 kHz output spread across a lot was ±3 dB. Two causes, both integration. The grille open area was about 2%, costing an estimated 4 dB across the 2–4 kHz band; and the hand-placed bead varied in compression by roughly ±0.15 mm, shifting F0 by several Hz unit to unit. The driver was not changed. The grille was re-cut to 5.5% open area, which recovered about 3 dB in the speech band, and the bead was replaced by a die-cut self-adhesive gasket of fixed 0.3 mm thickness located by the driver flange. After the change, 1 m intelligibility rose to 74 dB(A) and the 3 kHz spread tightened to ±1 dB across the same lot size. The lesson is that the two cheapest parts in the bill of materials — the grille pattern and the gasket — were the ones deciding whether the product passed.

The driver was fine. The grille and the gasket were the product.

  9. FAQ — Internal Speaker Integration

How much back volume do I actually need?

For a voice-prompt product, 0.5–2 cc of sealed cavity behind the driver is enough to define F0. For media reach you want 2–6 cc, and for real bass below 300 Hz you need 6–20 cc or a ported design. These are starting points — confirm against the actual enclosure measurement.

How do I size the grille open area?

Aim for at least 5% open area for any product that must be intelligible beyond a quiet room; below 2% you will lose 4–6 dB in the 2–4 kHz speech band. Design the hole pattern and any mesh behind it together, because a fine mesh compounds the loss.

What gasket approach is best for production?

A die-cut self-adhesive gasket of a fixed 0.2–0.4 mm thickness, located by the driver flange, gives roughly ±0.5 dB F0 spread across a lot and removes the operator variable of hand-beading. Reserve hand-beading for prototypes.

How do I stop a buzzing internal speaker?

Find whether the housing wall is flexing against the driver frame (stiffen it, add a standoff, or move to a box module), whether screw torque is uneven (specify torque and thread-lock), or whether the diaphragm is touching the grille (increase the standoff 0.5–1.0 mm). Buzz is mechanical, so the fix is mechanical.

Does the housing really shift F0 that much?

Yes. Roughly, doubling the sealed back volume lowers F0 by about 15–20%, halving it raises F0 by 15–30%, and an air leak raises F0 while dropping output below it. Always measure the actual enclosure before comparing to the catalog F0.

Should I seal my own enclosure or buy a box module?

If you have an acoustic engineer and a repeatable enclosure, sealing your own is cheaper. If you do not control the housing, or the launch date leaves no time to tune, a box module carries its own sealed volume and removes the housing from the acoustic load — which is also the most reliable cure for buzz.

Which adhesive and magnet grade for high temperature?

For outdoor or in-vehicle products from -10 to 70 °C, specify a high-temperature center adhesive (grades such as 4505, 868K or AB, chosen per program) and an H/M-grade magnet to hold F0 within about ±10 Hz and keep SPL up at temperature. Treat temperature range and THD as project-specific confirmations, not catalog defaults.

Can I pot the driver to stop buzz?

Potting can damp a voice-coil rub after shock, but it also changes the moving-system mass and can shift F0 and raise the resonance. Use it as a targeted fix for a specific failure, not a blanket cure, and re-measure after potting.

More in This Series — Internal Speakers for Consumer Electronics

This article is Part 2 of a three-part technical series on internal speakers for consumer electronics. Part 1 covers how to choose the part, and Part 3 covers symptom-level troubleshooting of finished devices.

· Part 1 — Selection Guide: Bare Driver vs. Box Module → https://www.hsdz-spk.com/news/517.html

· Part 3 — FAQ: Fixing Muffled Sound, Buzz and Weak Output → https://www.hsdz-spk.com/news/519.html

  10. Summary — Getting the Integration Right

Integration is where most internal-speaker programs win or lose. Four decisions decide the result: leave enough sealed back volume to set F0 where you want it, size the grille for at least 5% open area in the speech band, seal the back volume with a repeatable gasket rather than a hand bead, and mount the driver so the housing wall is not part of the moving system. None of these are the driver's responsibility, and none can be fixed later with DSP. For teams without an acoustic engineer, or on a fixed launch date, a box module that carries its own sealed enclosure removes three of the four decisions at once. When comparing options, look for a supplier that can tune the cavity and the gasket with you, not just ship a part with a datasheet.