Increasing bass damages the horn loudspeaker

The clipping lie

The most common technical nonsense?


"That wouldn't have happened to you with a larger power amplifier!" ...

This or similar advice is often given when someone asks "Why?" After a loudspeaker defect. represents.
Possibly once put into the world by an amplifier seller to promote sales, this claim is passed on today, especially in forums, as supposedly particularly intelligent "knowledge" and has not been killed for years. Yet it can be refuted by simple logical thinking!



Facts: what is clipping?
Clipping is the term used to describe the phenomena that occur when an amplifier is operated in an overdriven range. Then an amplifier can no longer reproduce the input signal amplified by a constant factor at the output, but has to "cut off" the largest peaks, which has an impact in the frequency range.
  • With 100% of the possible undistorted output power of an amplifier (often also referred to as nominal power or sinusoidal power), the amplifier can perfectly map an assumed sinusoidal signal from the input to the output. There are no interference products that would become noticeable as harmonics (= integer multiples of the input frequency) in the output signal. The distortion factor k as a measure of the ratio of harmonics to the fundamental in a signal is 0.0%.
    The power is 1,000 times the maximum output power, which is specified logarithmically as 0 dB (see the spectral line of the fundamental frequency n = 1).


  • If I just turn up the imaginary volume control, here, for example, so far that the sinusoidal voltage should be 1,286 times as high, the amplifier will distort. Cutting off the signal peaks creates harmonics, but only odd-numbered ones of the orders n = 3, n = 5, n = 7 etc. That means, I put a tone of 1 kHz on the amplifier, for example, but tones of 1 appear at the output kHz, 3 kHz, 5 kHz, 7 kHz, etc., which are not included in the input signal.
    However, their intensity decreases towards higher orders, see spectral lines that have been calculated here up to the 12th order. With this overload, the sum of the interference products in the overall signal results in the distortion factor k = 10%, the power is then 1,289 times the maximum output power.
    Since the distortion factor limit of 10% is often assumed to be a tolerable level for amplifier measurements, you can divide the specified amplifier outputs by 1,289 under these measurement conditions to arrive at the maximum undistorted sinusoidal output!


  • With even more overdrive, here up to a factor of 20, the distortion also increases further. With this, however, I can drive the distortion factor to a maximum of a little over 40%. The maximum possible power of an ideal square-wave signal is exactly twice the max.sine output power, i.e. a factor of 2.00!
    In practice, however, this would be a little less, since the power supply of an amplifier would sag a little due to the higher load.
    However, the fundamental wave it contains always remains the strongest component, but despite the overdrive factor of 20 only increases by a voltage factor of 1.27 (corresponds to 1.61 times the power or + 2.1dB). It is still 3.0 times as strong (i.e. 9.0 times the power) than the second strongest component, the third harmonic!




Consequences: What can I conclude from this?
  1. Every amplifier clips.
    A big one only from a high performance, but then with correspondingly large interference products.
    A small one from a small output, but then only with small interference products.

  2. For the consequences, it does not matter whether the clipping occurred in the output stage or is output in an upstream mixer or distortion for electric guitars and then output by a non-clipping amplifier. It is only relevant whether the power density existing in the frequency range concerned can be handled.

  3. Harmonics caused by clipping can kill a midrange or tweeter that otherwise would not have received any components from the original signal.

  4. The probability of this, however, is not very high, since the clipping harmonics have a significantly lower power intensity than a possible original wave or the music spectrum in this frequency range. This means that only mid-range speakers or tweeters can die through clipping, which would also die quickly without clipping, as they are simply dimensioned far too weakly for this amplifier!

  5. A bass or broadband can never be broken by clipping, because the clipping signal has overall less power than the non-clipping signal of a larger power amplifier would have. Even the effect of the harmonics can at most relieve rather than burden a bass, provided that the harmonics are outside its transmission range.

  6. If the amplifier gets into the clipping area and there is still an idiot at the controller who turns up even further despite the already audible distortion, the square-wave signal then gives twice the amplifier nominal output as continuous output.
    With a non-clipping larger amplifier, the possible overload is not even audible on the distorted signal and it doesn't even need an idiot to crank it up further ...
    In order to have security against burning through in the event of any incorrect setting, the amplifier output must therefore be dimensioned at least by a factor of 2 lower than the loudspeaker load capacity !!!

  7. Clipping also does not, as is often claimed, create direct voltage. The spectral line of the direct voltage with order n = 0 shows a deflection when there is no overload! The straight lines of a distorted signal do not represent a direct voltage, because they change polarity between + and - with the fundamental frequency, so they are alternating voltage!
    The straight sections also have no influence on the cooling of the voice coil: Because the membrane does not stay there for minutes, but blows a lot of air with a basic frequency between two extreme deflections!

  8. In this context, direct voltage can only arise as follows: If the positive limitation is different from the negative, e.g. at +51 volts and at -49 volts instead of symmetrically at + -50 volts, then this results in an average value of +1 for the extreme rectangle Volts DC voltage. Since even asymmetrically wound transformer windings alternately charge the plus and minus supply, this is unlikely and in any case small compared to the alternating voltages at the output. Such small performances do not harm any loudspeaker.

  9. Almost all loudspeakers that died (apart from the ones that were mechanically destroyed, e.g. by too much stroke, which has nothing to do with electrical power anyway) were not broken by clipping, but simply by too much electrical power.
    So not a larger but a smaller power amplifier would have helped!

  10. Never again believe anyone who advises you about clipping claims in forums or the like! Because he has come out as someone who has heard rumors and false statements without reflection, without even thinking about it!



Remedy: How can I really achieve security against burning through?
  • The real cause is always that the loudspeakers are too weak for the required volume (and the stupid attempt to compensate for this with power amplifier output)! Maybe you should just use more / louder boxes?
    A larger membrane area leads to more efficiency, especially in the bass (regardless of whether by using larger diameters or by interconnecting several smaller ones).

  • With more power than the loudspeaker can handle, it doesn't get louder, it distorts, heats and destroys even more! This is called compression (see image).

    Linear range: Due to the logarithmic relationship, the same power factors must always be added for the same dB intervals - in absolute differences, therefore, always greater watt intervals!
    Compression range: Here the loudspeaker is no longer louder in proportion to the increase in power. The causes are mechanical limitation and temperature-dependent increasing resistance. The additional power is used extremely ineffectively! Compression in dB is the distance between the real and the theoretical volume.
    Defect area: Can either only be reached for milliseconds or forever!

  • "But my amplifier wasn't turned up all the way."
    The position of the controller says nothing about the power, but only about the gain factor, i.e. the ratio of output level / input level! Even turned up at 1/10, 100% power is achieved at the corresponding input level!
    So this method does not provide absolute security.

  • "But I want a big power amplifier. Is there another way to protect the speakers from overload?"
    Possibly you can also add limiters, switch-off units or PTCs. But smaller amplifiers are certainly more intelligent than buying large power and preventing it from occurring again! PTCs are only used sensibly if they limit the power distributed in each case in two branches in the crossover.
    What an absurd thought to want to invest in performance in order to cut it off before it occurs!?!

  • "Then I set my limiter to exactly 99% of the speaker capacity!"
    There is no unambiguous number of resilience. This is related to the ambient temperature, frequency range, signal shape, impedances, duration, etc. In order to get burn-through protection, a limiter would have to be set in addition to the square / sine factor of 2.0 (see above) for reusable boxes by a factor of 2 to 5 below, since the power is not necessarily distributed, but momentarily 100% can also reach the tweeter (e.g. with feedback !!)!
    Any considerations of accuracy are out of place here, since the performance specifications from the manufacturer are no more than a rough guide (and often are also a lie!) And depending on the situation, vastly different performances lead to burnout.

  • "How do I recognize a credibly high speaker capacity?"
    There may be people who believe the computer boxes for 19.95 EUR with the 300 watts printed on them. But imagine a lightbulb in which the wire glows at 40 watts at over 1000 degrees! And you want to give hundreds of watts to a loudspeaker voice coil of comparable dimensions, whose insulating varnish can withstand a maximum of 150 degrees, for example?
    The most important thing is therefore a large voice coil diameter so that the heat is distributed. E.g. a 35mm voice coil with a nominal power rating of 200 watts is simply not credible, but is often stated that way. With the same diameter, a long-throw coil is then superior to a short mid-range coil. Ferrofluid also increases the load capacity enormously through better heat dissipation.

  • "How do I upgrade meaningfully if I don't want to spend a lot all at once?"
    The correct sequence of steps is important: First upgrade the speakers and only then the amplifiers (green path). Customers who did it the other way around (red path) often got stuck in step 2 and repeatedly had to spend money on replacement speakers.
    Within the loudspeaker combination, for example, you can first operate the later top parts in full range and then add subwoofers in the second step.

  • I also don't know why it is so difficult to just take a 100W power amplifier if the combination is to be safe. There are loudspeakers that are just as loud at 100 W as others only at 400 W. Does it always have to be boasted with a high wattage?