logo elektroda
logo elektroda
X
logo elektroda

Class D PAM8610 15V 2x10W amplifier module test

TechEkspert 15492 34

TL;DR

  • The PAM8610 class D 2x10W amplifier module adds a power connector, speaker terminals, a minijack input, a volume potentiometer, and a small stick-on heat sink.
  • Powered at 15V with 8 ohm loads, the module showed louder idle noise than PAM8403, and the output waveform looked unlike a typical class D amplifier.
  • With no input, current draw was 26mA unloaded, 30mA with 8ohm speakers, 41mA with 8ohm resistors, and 44mA with 4ohm resistors.
  • At 2.8V on 8 ohms, it delivered 2x1W with 0.423A input current; at 8.9V, it reached 2x10W and about 77% efficiency.
  • The module still played music acceptably, but it ran very warm at higher power and used two jumpers for power-on and mute.
ADVERTISEMENT
Treść została przetłumaczona polish » english Zobacz oryginalną wersję tematu
📢 Listen (AI):
  • #31 17258034
    irek2
    Level 40  
    The PAM8610 on the black board has a very high gain, as I remember it well, as much as 36dB. The signal level at the headphone output is very low and restricted by standards (hearing protection). Additional preamplifier needed.

    As for the shutdown, I suspect you had a problem with the power supply. Either too low capacity or too long connection cables and no additional electrolyte at the module.

    I bought two other modules for TPA3110 and OEP30, both of them have classic bipolar modulation, so definitely lower distortions and certainly much less noise, although the amplification is slightly lower.
  • ADVERTISEMENT
  • #32 17258108
    Łukaszoo
    Level 16  
    Quote:
    The signal level at the headphone output is very low and restricted by standards (hearing protection).

    I know, and it's fucking annoying. For comparison - I have an old thinkpad, there you can lead the tip to the clip from the headphone output. This is impossible with a modern telephone.

    But back to the PAM8610 case - I exclude the weak power supply, I tested it on several different power supplies, and now there is Aku 12V 18Ah.
    I also tested TPA3110 (white module) and nothing like that happened there. But the one without the preamplifier is completely useless.
  • ADVERTISEMENT
  • #33 17421598
    Łukaszoo
    Level 16  
    I am after quite a long fight with the module on PAM8610.
    In the final phase, such a "monster" with the capacity that the hands drop (2x1000u, 2200u, 220u L-ESR, 220n, 5x100n, 68n, 220p) was already soldered to the amplifier pins.
    and still had a problem with turning off. I made a simple "workaround" of the problem by passing the signal through a high-pass filter that cuts low frequencies to reduce the power consumption and ... here, so to speak, "the stick overdriven" - the amplifier started to shoot strangely, crackle, it had little to do with analog amplifier clipping. The loudspeaker started to "ripple" as if it were getting a few Hz with a considerable amplitude.
    I ran out of patience with this chip, TPA3110 took its place. The shutdown problem is gone, it distorts like analog, it's much better overall.
    The only thing that I pay attention to is much lower gain. Unfortunately, you have to play with the preamplifier. Fortunately, the usual, simplest, unbalanced-powered WO amplifier does the trick.
  • ADVERTISEMENT
  • #34 17558733
    Tremolo
    Level 43  
    To the author of the TechRedaktor topic. In the application, I am writing about adding the output inductance in the case of too short cables. The sound is good on long, thick cables.
    The circuit "pulls" in any distortion of the power supply on the switching power supply. There is no such problem on a weaker classic transformer power supply. In my view, it is used for car radios.

    Łukaszoo - what kind of cooling did you use? For me, the included microradiator is too weak.
    Helpful post? Buy me a coffee.
  • #35 17559815
    Łukaszoo
    Level 16  
    I have made various tests with cooling.

    Without a heat sink, with the factory (attached to the module), with a slightly larger frame removed from the bone, with two at the top and bottom, I even had to pour isopropanol on it while working. No effect on the behavior of the system, so I ruled out overheating. And I tested three such modules, so the fact that I got a faulty one is also unlikely.
    It's great for a small portable speaker, but for heavier plowing, I recommend TPA. Apart from too low gain, it did not cause me any problems with the application.
📢 Listen (AI):

Topic summary

✨ The discussion revolves around the PAM8610 Class D amplifier module, highlighting its features, performance, and testing experiences. Users note its compact size, affordability, and suitability for portable audio applications, particularly with battery power. However, concerns arise regarding sound quality, efficiency, and operational issues, such as noise generation and power supply problems. Comparisons are made with the PAM8403 module, with some users reporting better performance from alternative amplifiers like TPA3110. Suggestions for improving performance include adding output filters to reduce noise and distortion. The conversation also touches on the theoretical efficiency of Class D amplifiers versus Class AB, emphasizing practical limitations in real-world applications.

FAQ

TL;DR: A PAM8610 board delivered 2 × 10 W at 77 % efficiency, yet “noise louder than PAM8403” [Elektroda, TechEkspert, post #17198703] Good for battery speakers when fed with a stout 12 V source and short leads. Why it matters: you can build loud, long-lasting portables without bulky heat-sinks.

Quick Facts

• Operating voltage: 7–15 V DC (abs. max = 16.5 V) [Elektroda, irek2, post #17206215] • Peak power: 2 × 15 W into 4 Ω at 12 V, 1 kHz, ≤10 % THD [PAM8610 Datasheet] • Idle current: 26–44 mA depending on load [Elektroda, TechEkspert, post #17198703] • Board gain: 32–36 dB (≈40×–63×) [Elektroda, irek2, #17206215; #17258034] • Typical AliExpress price: US $1.50–$3 (2024) [AliExpress Search]

1. What continuous power and efficiency can I really expect?

With 15 V supply and an 8 Ω load, testers measured 2 × 10 W RMS and 77 % efficiency before the IC became “very warm” [Elektroda, TechEkspert, post #17198703] Lower levels—2 × 5 W—came in at ~61 % efficiency.

3. Why do I hear hiss or "chaff" with no signal?

Without input, the modulator still switches, and the missing LC filter lets carrier residue reach the cone, creating audible hiss [Elektroda, mkpl, post #17198820] Differential inputs left unused can also pick up RF dirt.

4. How can I cut switching noise and FM interference?

  1. Keep speaker leads under 20 cm.
  2. Add an LC π-filter (e.g., 10 µH + 220 nF) per channel.
  3. Shield the board in a metal enclosure tied to ground. Users report VHF noise vanishes with these steps [Elektroda, mkpl, #17198820; kris1966, #17217953].

5. The board shuts off at high volume—what fixes it?

Voltage dips trigger over-current protection. Use a supply rated ≥12 V 5 A or 3-cell Li-ion pack, add a low-ESR 1000 µF capacitor at the pins, and keep leads short. One test showed >5 A current spikes during music [Elektroda, irek2, post #17206215]

6. Do I need a pre-amp for phones?

Probably yes. Board gain is 32–36 dB, yet many modern phones cap headphone output at ~100 mVp. Users noted some phones could not reach full loudness without an op-amp stage [Elektroda, Łukaszoo, post #17254743]

7. What cooling is required?

At 2 × 10 W the PAM8610 gets hot but survives with a 20 × 20 mm stick-on sink and airflow. Tests showed shutdowns were supply-related, not thermal—even with no sink [Elektroda, Łukaszoo, post #17559815]

8. How do I probe outputs safely?

The outputs are bridge-tied. Float your scope or use 4 k7 Ω + 10 nF RC filters to ground on each lead, then sum channels with math mode. This reveals a clean sine without shorting the bridge [Elektroda, irek2, post #17206215]

9. Can I drive 4 Ω speakers?

Yes, but expect more heat and early clipping. A 12 V rail yields up to 2 × 15 W into 4 Ω, yet several builders saw the chip mute under heavy bass unless supply and caps were upgraded [Elektroda, Łukaszoo, post #17421598]

10. Quick 3-step LC filter build (How-To)

  1. Solder a 10 µH inductor in series with each speaker lead.
  2. Add a 220 nF film capacitor from lead to ground.
  3. Place parts within 2 cm of the PAM8610 pins. Result: carrier attenuation >30 dB and reduced idle hiss [Elektroda, mkpl, post #17198820]

12. Can the board’s gain be changed?

Some versions expose G0–G2 pins; desoldering jumpers lets you select gains from 20 dB to 32 dB per datasheet. Black no-pot boards are fixed at ~36 dB [Elektroda, irek2, post #17258034]

13. What edge-case failures should I know?

Wrong grounding or long input leads can cause oscillation, smoke, and IC destruction, as one user found after random shutdowns and a flash fire [Elektroda, kris1966, post #17202929]

14. Is it suitable for car audio?

Yes, if you add robust filtering. The chip tracks battery voltage and rejects alternator ripple, but switching noise can enter FM tuners unless the module is screened and output leads filtered [Elektroda, Tremolo, post #17558733]
ADVERTISEMENT