Small Speaker Driver FAQ: Burnt Voice Coils, Amplifier Shutdown and Lost Volume

Writer:By Shenzhen Hongsheng Electronic Industry Co. LTD Visits: 09 07, 2026

Small Speaker Driver FAQ: Burnt Voice Coils, Amplifier Shutdown and Lost Volume

Published: 2026-09-04  |  Troubleshooting scope: fault diagnosis at the driver and amplifier interface in smart home hardware — smart door locks, video doorbells, wall panels, hubs, thermostats and appliance control panels — covering burnt coils, thermal shutdown, missing output, distortion and impedance mismatch.

When audio fails in a smart home device, the driver is usually blamed and usually innocent. Burnt voice coils, sudden volume loss and output that never matched the datasheet almost always trace back to the electrical interface: a load the amplifier was never rated for, a tone held at full gain, or a sensitivity figure quoted on a measurement basis the product does not reproduce. Take Shenzhen Hongsheng Electronic Industry Co. LTD as an example: in the field-return investigations we support, measuring the DC resistance of the suspect coil and the RMS voltage across the driver terminals identifies the cause within minutes, long before acoustics enters the conversation. This guide maps symptoms to electrical causes, gives the measurements that settle each one, and lists the production drivers suited to each fault signature.

  1. Symptom Map for Driver-Level Failures

Table 1: Symptom to cause map, ordered by how often each appears in field returns

Symptom

Most likely electrical cause

First measurement

Typical verdict

No sound at all, driver reads open circuit

Voice coil burnt through from sustained overdrive or a clipped amplifier

DC resistance across the terminals

Infinite resistance — coil is open, driver must be replaced

Volume collapses after 30-60 seconds of loud playback, recovers when cold

Amplifier thermal or overcurrent foldback, usually from too low a load impedance

Load impedance against the amplifier's rated load; case temperature of the amplifier

Amplifier is protecting itself; the driver may be undamaged

Output well below the datasheet figure, clean sound

The amplifier cannot deliver the power the sensitivity figure assumed, or the figure was quoted on a different basis

RMS voltage across the driver terminals during a prompt; compute V squared divided by impedance

Specification mismatch, not a fault

Distortion at high level, harshness on peaks

Amplifier clipping, or the voice coil leaving its linear range

Waveform at the driver terminals; check for flat-topping

Gain structure problem or excursion limit reached

Buzz or rattle that appears on some units only

Mechanical: voice coil rub after shock, loose particle in the gap, or a resonating panel

Tap test and a swept sine; DCR stays normal

Mechanical, not electrical

Hiss or hum with no signal present

Amplifier noise floor, layout or supply decoupling; not a driver fault

Noise with the driver disconnected

Source-side problem

Intermittent sound, comes and goes with temperature

Fatigued lead-wire joint or a cracked solder termination

Resistance while flexing the leads gently

Termination defect

  2. Burnt or Open Voice Coil

A burnt coil is a thermal event, and it always has an energy source. The three sources that account for nearly every case are sustained drive above rated power, a clipped amplifier stage, and a load impedance the amplifier was never rated to feed. The third is the one most often introduced late in a program, when a driver is swapped for a better sensitivity figure without rechecking the electrical interface.

Clipping deserves particular attention because it is invisible on a specification sheet. An amplifier driven past its clean limit stops producing a larger sine and starts producing something closer to a square wave, and the average power in that waveform is far higher than the rating suggests. Driving a stage about 6 dB past its clean limit can put roughly four times its rated power into the load, and the voice coil has to dissipate essentially all of it as heat.

Table 2: Overdrive mechanisms, the evidence each leaves, and the fix

Mechanism

How the failure is introduced

Evidence

Fix

Sustained drive above rated power

An alarm chime or alert tone held at full gain for longer than the coil can dissipate

Open coil; discolouration at the former; no amplifier fault

Duty-cycle the tone and size the driver one power class up

Clipped amplifier output

Digital gain set too high, so routine content drives the stage past its clean limit

Open or high-resistance coil; the amplifier itself tests normal

Cap the digital gain 3-6 dB below full scale and re-measure

Load below the amplifier's rated impedance

A 4 ohm driver fitted to a board laid out for 8 ohms

Open coil, often with a failed or degraded amplifier alongside it

Return to the rated load, or change the amplifier to a 4 ohm stable part

Blocked thermal path

Driver potted, gasketed or enclosed so coil heat cannot leave

Open coil with no electrical overdrive evident

Provide a vent path or reduce the duty cycle

The measurement that settles it takes seconds. Read the DC resistance across the driver terminals and compare it with the nominal impedance. A 4 ohm driver typically measures about 3.2 to 3.6 ohms, and an 8 ohm driver about 6.5 to 7.2 ohms, because DC resistance sits roughly 15 to 20 percent below nominal impedance. An infinite reading means the coil is open. A reading about 20 percent low points to shorted turns, which is a partially burnt coil that may still pass a quick bench test and will fail in the field.

  3. Amplifier Shutdown, Foldback and Clipping

Thermal and overcurrent protection exist to save the amplifier, and when they engage the symptom is a specific one: the device works when cold, loses volume or mutes entirely after half a minute or so of loud output, and recovers fully once it has cooled. That pattern identifies the amplifier as the protecting party rather than the failing one, and it means the driver may be undamaged.

Table 3: What to check when protection engages

Check

How to perform it

What a pass looks like

If it fails

Load impedance versus amplifier rating

Read the driver's nominal impedance against the amplifier datasheet's rated load

The load is at or above the rated minimum

Move to a higher-impedance driver or a lower-impedance-capable amplifier

Current demand at the rail

Divide the rail voltage by the load impedance for the peak current estimate

Peak current is within the amplifier's limit with margin

Reduce gain or raise the load impedance

Thermal pad and copper area

Inspect the amplifier's exposed pad solder coverage and the copper area beneath it

Pad fully soldered with the copper the datasheet specifies

Add copper, vias, or both; check the layout against the reference design

Supply decoupling and droop

Scope the rail during a loud prompt

Rail holds within a few percent

Add bulk capacitance or shorten the supply path

Gain structure

Measure the waveform at the driver terminals at maximum volume

Clean sine with no flat-topping

Reduce digital gain until the peaks are clean

  4. Output Lower Than the Datasheet Promised

This is the most common complaint and the least likely to be a defect. The published sensitivity figure is measured at a stated power, at 10 cm, at a stated frequency, in a stated enclosure. In the product, one or more of those conditions differs, and the shortfall is arithmetic rather than failure.

Table 4: Where the missing decibels actually go

Cause

Typical size of the loss

How to confirm

What to do

Amplifier cannot reach the driver's rated power at the rail

1-3 dB

Measure RMS voltage across the terminals; compute V squared over Z, and compare with the rating

Move to the impedance class the amplifier can drive, or raise the rail

Sensitivity quoted at 2.83 V rather than 1 W on a 4 ohm part

About 3 dB of apparent advantage

Read the measurement basis on the datasheet

Convert both candidates to the same basis before comparing

Distance: datasheet is at 10 cm, requirement is at 1 m or 2 m

20 dB at 1 m, about 26 dB at 2 m

Identify the measurement distance on the datasheet

Budget for it during selection, not after

Restrictive grille

2-6 dB, concentrated in the 2-4 kHz speech band

Measure with and without the grille fitted

Increase open area towards 5 percent of the driver face

Leaking cavity or compressed gasket

2-4 dB of low-frequency loss

Compare assembled against a known-good seal

Use a die-cut fixed-thickness gasket and state a leak rate

Firmware limiter or conservative gain

3-10 dB

Check the DSP gain and limiter settings

Raise gain only after the clipping margin is confirmed

  5. Impedance Mismatch and the Wrong Load

The consequences of a mismatch are strongly asymmetric, and knowing which direction is dangerous is most of the diagnostic value.

Table 5: Mismatch direction, consequence and severity

Situation

What the amplifier does

Consequence

Severity

8 ohm driver on a 4 ohm-capable amplifier

Delivers about half the power it is capable of

Roughly 3 dB of output lost

Wasteful but safe

4 ohm driver on an 8 ohm-rated amplifier

Asked for twice the current it was designed to supply

Overcurrent or thermal foldback during loud passages; possible amplifier damage

Potentially destructive

Two drivers wired in parallel

Halves the combined load impedance

Same risk as the row above; a second 8 ohm driver in parallel presents 4 ohms

Destructive if not designed for

Two drivers wired in series

Doubles the load impedance

Safe, but each driver receives a quarter of the power a single one would

Safe, and usually disappointing

32 ohm receiver part on a loudspeaker output

Very little current flows

Low output, and the part is on a coupler measurement scale anyway

Not a defect, but not a workable substitution

  6. Distortion and Rubbing at High Level

Distortion at high level has two very different origins, and they are separated by a single measurement. If the waveform at the driver terminals shows flat-topping, the amplifier is clipping and the driver is innocent. If the waveform is clean and the sound is still harsh, the driver has run out of excursion or is rubbing mechanically.

Table 6: Mechanical faults and how to tell them apart

Fault

How it sounds

Confirming test

Remedy

Voice coil rub after mechanical shock

Intermittent buzz that changes when the housing is pressed

Swept sine; the buzz appears at particular frequencies and varies with orientation

A spider-located or better-damped construction improves shock tolerance

Excursion beyond the linear range

Harshness that appears only at high level and disappears 3 dB lower

Reduce level by 3 dB; if the harshness vanishes, it is excursion

Lower the gain, add a high-pass filter, or move to a larger driver

Loose particle in the magnetic gap

Scratchy, level-independent noise present on all content

Tap the frame lightly while playing a tone

Replace the driver; do not attempt to clear it

Panel or grille resonance

Buzz at specific notes, absent when the housing is held firmly

Press the suspect panel while sweeping

Stiffen the panel, add damping, or change the mounting

Excursion data is scarce for parts of this size. Only one driver in this catalog publishes an Xmax figure — HS001846H at 0.8 mm — so for every other part the excursion ceiling has to be established by measurement on a swept sine rather than read from a datasheet.

  7. FAQ — Small Speaker Driver Troubleshooting

How do I tell a burnt voice coil from a faulty amplifier?

Measure the DC resistance across the driver terminals. An infinite reading means the coil is open and the driver is dead. A reading in the expected range — roughly 15 to 20 percent below the nominal impedance — means the driver is electrically intact and the fault is upstream, most likely amplifier protection engaging. If the coil is open, also check that the load impedance matches what the amplifier is rated for, because a low load is a common cause of the burnout rather than a consequence of it.

Why does my device lose volume after a minute of loud playback and recover later?

That is amplifier thermal or overcurrent protection, and the driver is probably fine. The usual cause is a load impedance below what the amplifier was rated for: a 4 ohm driver on a stage designed for 8 ohms asks for twice the intended current. Check the rated load in the amplifier datasheet, then either raise the load impedance, reduce gain, or improve the thermal path under the amplifier.

The driver is rated 97 dB but the product measures far less. Is it faulty?

Almost certainly not. Convert the figure before judging it. A 97 dB rating at 10 cm is about 77 dB at one metre and about 71 dB at two metres, because every doubling of distance costs 6 dB. Then check what power the amplifier actually delivers: measure the RMS voltage across the terminals and divide its square by the impedance. If that is below the driver's rated power, the shortfall is arithmetic, not a defect.

Can I replace an 8 ohm driver with a 4 ohm one for more output?

Only if the amplifier is specified as 4 ohm stable and the supply can provide the extra current. On a 5 V bridge-tied-load stage the change is worth roughly 3 dB, which is real, but it doubles the current demand and can trigger protection or damage the amplifier if it was not designed for it. On battery-powered devices the extra current also shortens run time, so the trade is usually poor.

What DC resistance should I expect from a 4 ohm or 8 ohm driver?

Roughly 15 to 20 percent below the nominal impedance, because DC resistance is not the same quantity as rated impedance. A 4 ohm driver typically measures about 3.2 to 3.6 ohms and an 8 ohm driver about 6.5 to 7.2 ohms. An infinite reading indicates an open coil. A reading around 20 percent lower than expected indicates shorted turns, which is a partially burnt coil that may still pass a brief bench test.

How much power can I safely put through a driver rated 1.0 W?

It depends entirely on the duty cycle. Rated power is a continuous thermal limit under the standard test signal, so for a continuous alarm tone you should stay at or below the rated figure. For short voice prompts with a low duty cycle, brief peaks above the rating are normally acceptable, because what matters is the RMS power over a prompt cycle rather than the instantaneous peak. If the design runs close to the limit, specify a lead-wire or spider-located construction, which are the two upgrades that reliably improve power durability.

Is distortion at high volume a driver problem?

Check the waveform before replacing anything. Flat-topping at the driver terminals means the amplifier is clipping, and reducing digital gain 3 to 6 dB will fix it. A clean waveform with harsh sound means the driver has reached its excursion limit or is rubbing mechanically. Dropping the level by 3 dB is a quick discriminator: if the harshness disappears, it was excursion.

Can I wire two drivers in parallel for more output?

Only with the electrical consequence in mind. Two 8 ohm drivers in parallel present 4 ohms to the amplifier, which is fine if the stage is 4 ohm stable and otherwise destructive. Note also that paralleling shares the power between the drivers rather than doubling it, so the acoustic gain is smaller than it looks — roughly 3 dB from the halved load, partly offset by each driver receiving half the power. Wiring in series is always safe and usually disappointing.

Where can I find an excursion limit for a driver this small?

Usually nowhere on the datasheet. Across this catalog only one part publishes Xmax — HS001846H at 0.8 mm. For everything else, establish the limit by measurement: drive a swept sine and raise the level until distortion exceeds an acceptable threshold, then set the operating ceiling a few decibels below that point. Treat any published excursion figure for other parts as an estimate rather than a specification.

  8. Step-by-Step Diagnostic Checklist

1. Confirm the complaint with a measurement, not an impression. Note the distance, the content and the ambient temperature at which the fault appears.

2. Measure the DC resistance at the driver terminals. Infinite means an open coil; about 20 percent low means shorted turns; in range means the driver is electrically intact.

3. Read the load impedance against the amplifier's rated minimum load. A driver below that minimum is the single most common root cause of burnt coils and thermal shutdown.

4. Measure the RMS voltage across the driver terminals during the failing content, and compute the delivered power as voltage squared divided by impedance. Compare it with the driver's rated power.

5. Scope the waveform at the same time. Flat-topping means clipping, and the fix is gain structure rather than a new driver.

6. Check the amplifier's thermal path: exposed-pad solder coverage, copper area and vias, and whether the rail holds under load.

7. If the driver is intact and the power is correct, look for acoustic causes: grille open area, gasket compression, cavity leaks and panel resonance.

8. For intermittent faults, flex the lead wires and the terminals while monitoring resistance, then repeat the test at the temperature extremes the device is rated for.

9. Only after all of the above, substitute a known-good driver — and record the DC resistance and delivered power of both units so the comparison means something.

Table 7: Production drivers suited to each fault signature

Fault signature

What the replacement needs

Candidate models

Coil burnt by a sustained alarm tone

Higher power handling and better thermal path

HS284011H (3.0/4.0 W, spider), HS003050H50 (2.5/3.0 W, lead-wire voice coil)

Thermal shutdown on a 4 ohm load

A higher impedance so the amplifier is within its rating

HS003050H or HS003650H (8 ohm, 2.0/2.5 W, 97 dB)

Output short at the same impedance

More sensitivity with no change to the electrical load

HS241534H34 (dual magnet, 95 dB at 1.0 W), HS004550H (98 dB)

Excursion-limited harshness at high level

A larger diaphragm or a suspension-located coil

HS284011H (28 x 40 x 11, spider), HS0028110H110 (foam edge)

Rubbing after shock or drop

A spider-located or better-damped construction

HS284011H (spider), HS0034140H140 (pot-type large magnet)

Low-frequency body missing from prompts

Lower F0 from a pot-type large-magnet circuit

HS0034140H140 (300 Hz), HS0028110H110 (350 Hz), HS0023123H123 (400 Hz)

Cavity unavailable or unreliable

An integrated box module instead of a bare driver

HS003021H (105 dB), HS003058H (103 dB), HS002628H28 (99 dB)

Battery device needing output without more current

More motor rather than a lower impedance

HS241534H34 (dual magnet, +2 dB at 8 ohms and 0.6 mm thinner)

Project Case Study — Burnt Coils in a Video Doorbell

A video doorbell deployed in a warm climate began returning units with silent audio after a few months in service. Teardown showed open voice coils, and the obvious hypothesis was a defective batch of drivers. The measurement that redirected the investigation was the load: the board carried a Class-D stage specified for 8 ohms, but the driver fitted was a 4 ohm part rated 2.0 W, substituted late in the program because its published sensitivity was 1 dB better. At the 5 V rail that amplifier was being asked for roughly twice the current it was designed to supply, and the doorbell's chime — a continuous tone held at maximum gain for several seconds — drove it into clipping, which put approximately four times the rated power into the coil. Three changes resolved it: reverting to the 8 ohm driver in the same 30 mm outline, duty-cycling the chime to 40 percent, and capping the digital gain 4 dB below full scale. Shenzhen Hongsheng Electronic Industry Co. LTD verified the revised build against an IEC 60068-2-2 dry-heat soak with the coil temperature monitored and the chime repeated at the maximum duty the firmware permits. Field failures for audio fell from 3.1 percent to under 0.2 percent across the following two quarters, and the 1 dB of published sensitivity that had started the problem was never recovered — nor missed.

More in This Series — Small Speaker Drivers for Smart Home Devices

This article is Part 3 of a three-part technical series on small speaker drivers for smart home devices. The other two parts cover how to select a driver against a fixed amplifier, and the impedance and power matching in engineering detail.

· Part 1 — Selection Guide: Choosing a Driver for a Fixed Amplifier → https://www.hsdz-spk.com/news/526.html

· Part 2 — Technical Requirements: Impedance, Power Handling and Amplifier Matching → https://www.hsdz-spk.com/news/527.html

  9. Summary — Keeping Drivers Alive

Diagnose at the interface before replacing the part. A DC resistance reading separates an open coil from an intact driver in seconds, and an RMS voltage measurement across the terminals tells you whether the amplifier is delivering what the datasheet assumed. Most field failures in this class trace to three causes: a load below the amplifier's rated minimum, a tone or chime held at full gain, and a gain structure that clips routinely. All three are fixed in firmware, layout or specification rather than in the driver. When a replacement is genuinely needed, choose it against the same electrical interface the original was meant to see, and favour lead-wire or spider-located construction wherever the design will run near its thermal limit.