Και πριν μου τα σούρει κάποιος,ας διαβάσει το παρακάτω:
1.8 - Passive Crossovers
For those who are unfamiliar with the setup of a three-way passive crossover, please refer to Figure 5, which shows (and the text explains) the connections. The diagram shown is for a "bi-wired" system, but includes the conventional connections.
When an amplifier reproduces the entire musical range, coils (inductors) and capacitors are used in the speaker cabinet to separate the high and low frequencies so that each may be supplied to the appropriate loudspeaker driver. A loudspeaker can be a difficult load for any amplifier, but when additional inductance and capacitance enter the equation, this only makes matters worse. Add to this the fact that all passive crossovers introduce some degree of loss (in some cases as much as 3dB - which means that they are "stealing" half the available power), and one can see that getting rid of them cannot be such a bad thing.
Look at the impedance graphs for almost any speaker system, and it will be seen that there is almost always a dip in impedance (sometimes severe) at the crossover frequency. This is caused by the interactions of the loudspeakers and their inductor/capacitor networks, and in some cases can cause amplifiers considerable grief - especially at high power levels. Although few amps will fail, one can expect a reduction in effective output power as the protection circuits limit the maximum power available due to the loading of the crossover network.
These vague thoughts are brought into stark reality when one learns that the inductors and capacitors needed for the low frequency crossover are quite large values, which leads many speaker designers to compromise in the interests of economy. The inductors may have an iron or ferrite core - which improves its inductance, but ruins its linearity. So now the crossover behaves differently depending upon the amplitude of the signal. High value high quality capacitors are expensive, so again, bi-polar electrolytics are often used. It is often stated that these sound awful, although this is a somewhat contentious issue, but without any doubt their characteristics change with temperature and age. They also have rather mediocre accuracy against their claimed value (+20/-50% is typical), so a 10uF crossover cap may be 12uF, or as low as 5uF. This does not make for an accurate crossover network, and most reputable speaker manufacturers will not make this sort of compromise, at least not for their top-of-the-line models.
In addition, at high powers, the impedance of the voice coil rises because of the temperature rise in the voice coil. This is not stable, and varies widely with the music. So with loud passages, the voice coil temperature might rise significantly, which will severely impact the performance of the crossover - relying as it does on the load impedance being a constant. A loud bass solo followed by a relatively quiet but complex passage might create an interesting shift in the crossover frequency and phase response as the voice coil cools, which is unlikely to enhance the listening experience.
Κι αν θέλει όλη την ανάλυση,
εδώ.
